Chapter 72. Management of Acute Postoperative Pain

1. Analgesia, as distinct from hypnosis, is a vital and integral component of anesthesia.

2. Anesthesiologists must plan for the continuum of intra- and postoperative pain.

3. The principles of "opioid sparing" or "multimodal analgesia" are central to the goal of rapid recovery because opioid side effects delay recovery.

4. Patient-controlled analgesia (PCA) has greatly facilitated acute pain management at both provider and institutional levels.

5. Epidural analgesia continues to play an important role in the treatment of pain after major intra-abdominal and thoracic surgery, although the benefit versus risk should be reexamined in an era of potent thrombosis prophylaxis.

6. Chronic opioid use and abuse are emerging as prominent challenges during acute pain treatment.

After the first public demonstration of ether anesthesia at the Massachusetts General Hospital in 1846, the news that "We have conquered pain" spread around the world. What was not understood at the time is that the potent hypnotics merely suspend pain perception but do little to change pain transmission. The addition of neural blockade or strong analgesics (notably opioids) is needed to halt the pain processes that excite the nervous system, trigger neuroendocrine stress responses, and produce pain once it can be perceived. The provision of analgesia, during and after surgery, is now considered a vital and integral part of anesthesia, and its central goal to reduce the stress and derangements of surgery. Anesthesiologists apply their knowledge of pharmacology, anatomy, physiology, pathophysiology, surgery, and medicine toward optimizing pain relief. Whether by means of an informal relationship with surgical colleagues or a more structured service—an acute pain service—they play a key role in managing postoperative pain.1 As experts, they help educate others in the tools of pain management and teach the importance of pain control in terms of postoperative recovery.

Institutional Considerations

The Role of the Acute Pain Service

The idea of the acute pain service arose during the 1980s when "walking" epidurals made epidural analgesia suddenly more feasible for postoperative patients and when the microchip made patient-controlled analgesia (PCA) pumps small enough to have wide applicability. Initially, pain services ran both these modalities, developed treatment protocols, and taught nurses and others how to manage these new therapies. Soon, surgeons and nurses became familiar with the use of PCA, so that this component of postoperative management, at least for routine cases, is largely being managed by them. Epidurals remain the province of anesthesiologists, and the core function of the acute pain service is to manage postoperative epidural analgesia. The acute pain service is also available to help with complex cases, notably cases that cannot be managed using routine measures. Naturally, each institution will structure its pain service differently, according to institutional and local factors. Smaller hospitals and ambulatory facilities may not have a service as such.

Acute pain services are usually staffed by a mixture of attending physicians (sometimes pain trained and commonly anesthesiologists), nurses, physician assistants, pain fellows, and anesthesia residents. Rotation to the acute pain service is a valuable component of residency training because this is likely the only point during training that anesthesia residents have the opportunity to follow patients postoperatively on a daily basis. The allocation of roles on the service will depend to a large extent on factors such as reimbursement, hospital policies, the relationship between anesthesia and surgery, and the expectations and support of both the hospital and the department of surgery. For example, funding for nursing on the service varies according to institution, and the provision of nonbillable services, such as daily follow-up of PCA, may only be feasible if nurses are available and funded. Epidural management remains the most favorably reimbursed of acute pain services. Unfortunately, as pain management becomes increasingly challenging because of rising numbers of opioid-tolerant patients (both substance abusers and opioid-treated chronic pain patients), and because of mandated pain management, reimbursement often fails to cover the realistic costs of ideal pain management.

Pain Management and Quality Assurance

Pain management has become an important metric by which standards of medical care are measured, and there are sound reasons for this. First, experience tells us that despite the extensive efforts of advocates,2-4 pain is easy to neglect, it being "silent" unless it is sought. It becomes necessary, then, to find mechanisms for encouraging active pain management, knowing that educational efforts alone have limited impact. Although it is easy for busy practitioners to skirt over pain issues, pain is not something patients easily forget, and it is now recognized that there is a clear correlation between patients' satisfaction with pain management and their satisfaction with medical care overall.5,6 In a competitive health care milieu, pain management thus becomes important to hospital administrations.

The Joint Commission on Accreditation of Healthcare Organizations (JCAHO), known as The Joint Commission, became involved in pain management when it was recognized that published guidelines, including an acute pain guidelines published by government and professional bodies, had a limited impact on hospital practice.7,8 JCAHO evaluates and accredits more than 19,000 health care organizations in the United States and its status as an accreditation body is such that it serves as an alternative to state and federal inspection mechanisms. Because of its wide reach, it was chosen for collaboration when various bodies seeking to improve pain management—including the American Pain Society, the Agency for Healthcare Policy and Research (now the Agency for Healthcare Research and Quality), and the University of Wisconsin Medical School (supported by a grant from the Robert Wood Johnson Foundation)—determined that the accreditation body could potentially improve pain's visibility by mandating the development of processes that increase provider accountability for the assessment and treatment of pain. Five elements were established that were considered essential for establishing institutional responsibility for making pain management a priority9:

• Pain must be recognized and treated promptly.
• Information about analgesics should be readily available to clinicians in a way that facilitates order writing and interpretation of orders.
• Patients must be assured of attentive pain care.
• Explicit policies should be developed for the use of advanced analgesic technologies.
• Processes and outcomes should be examined for continued improvement.

The idea was to make pain management a priority in the health care system and to establish systems to support, reinforce, and reward good pain management practice. The standards were approved in 1999 and formally introduced into the accreditation process in 2001.10 Perhaps the most obvious manifestation of the mandate was the introduction of "pain as a fifth vital sign," part of the effort to "make pain visible." Many hospitals have now introduced a requirement to record a pain score on the vital signs chart. Although it was never the intention of JCAHO or the pain advocates that this should become the focus of the pain mandate, and indeed there is no written requirement for the "fifth vital sign" in current JCAHO documentation, nevertheless it has become the hallmark of the JCAHO mandate.

Measurement of patient satisfaction with medical care has become an essential part of hospital management in the competitive health care environment, and satisfaction with pain care is an integral component. Yet curiously, patients often express satisfaction with pain management even if conventional goals of pain care such as pain relief or improved mobility are not achieved.11,12 This paradox can probably be explained at least in part by the high patient satisfaction that arises by simply addressing pain.13 Pain relief scores per se seem less important than do the patient-clinician interactions, validation of the pain complaint, and confidence that relief is available.12 One must question also the validity of patient satisfaction measures. Patient satisfaction instruments have limitations that include lack of psychometric standards and reliance on surveys that tend to have poor reliability and validity.14,15 Patient satisfaction is a complex concept with sociodemographic, cognitive, and affective dimensions, in their turn influencing the expectation that underpins satisfaction.16 All in all, what seems important is that pain is addressed because the mere addressing of pain can result in less distress for patients, translated into improved satisfaction.

Although the efforts of accreditation bodies and hospital management systems toward improving pain care have undoubtedly had positive effects, there have also been unexpected adverse consequences, most related to opioid prescribing. Mandated pain management, even if it does not specify opioid use per se, inevitably encourages opioid prescribing. In the case of acute pain, the introduction of hospital-wide protocols that link pain scores to opioid dosing have resulted in a rash of opioid overdosing, sometimes with disastrous consequences.17 Clearly, opioids must always be prescribed on a case-by-case basis, and what is learned is that there is no easy substitute for patient-clinician interactions and continuous assessment and dose adjustments. Mandated pain management can sometimes bring clinicians into conflict with their patients when demands for opioid are seen as irrational or due to addiction. Patients have a stated right to receive opioids,2-4 yet experience shows that indiscriminate use of opioids can have disastrous consequences.18,19 Clinicians may feel pressured into prescribing against their best judgment not only because of mandates, but also because they may need to meet patient satisfaction metrics. Opioid prescribing must always be balanced, and because opioids can be dangerous if not prescribed judiciously, it is important that mandates and efforts to meet quality metrics do not interfere with sensible opioid use.

Preparation

Patient Education

Patients who are informed about their likely postoperative experience are much better able to cope with pain and the other discomforts of the postoperative state than those who enter the experience uninformed.20 Whether it is the surgeon, the anesthesiologist, or the nurse who talks to the patient about the postoperative course, the message should always be that the team will use all methods available to make the experience tolerable but that a goal of total pain abolition is unrealistic. This approach makes patients less fearful when they find that pain has not been completely removed. Anxious and fearful patients do not handle pain well, patients become fearful very quickly if they experience unexpected pain, and telling patients that the methods we have available are good enough to remove all pain is unhelpful. How often do we hear enthusiastic residents or nurses tell patients that if they accept an epidural, they will not have any pain?

Patients should be informed of their options with regard to pain management, preferably early in the preoperative course. The use of brochures or pamphlets describing analgesic options is helpful. A description of risks as well as benefits is necessary. Some patients expect to be informed and to take an active role in decisions regarding their care; others prefer that decisions be made by their medical team. In either case, informed patients are better prepared psychologically than those who have not received the benefit of being informed.

The preoperative visit is also an opportunity to teach patients how to communicate their pain. For example, if they are likely to be asked to relate pain using a verbal pain score, they should be taught about this. They should be encouraged to express pain and not try to be stoical. They should learn that uncontrolled pain may have undesirable consequences such as splinting of the diaphragm, atelectasis, and possible chest infection. They should learn how to use a PCA pump if they are likely to receive one. In particular, they should be aware that early on, they will need supplementary analgesia and should not expect that pressing the button will necessarily remove the severe pain that can occur in the immediate postoperative period. They should also be aware of the inherent safety of PCA and that overdose is unlikely.

Anesthesia Planning

Anesthesiologists can have a significant impact on patients' postoperative pain experience, even if they are only involved in intraoperative care. This is not only by means of sustainable analgesia (epidurals and other catheter treatments), but also by anesthesia planning that results in a pain-free or near pain-free emergence. The importance of a pain-free emergence cannot be overemphasized; patients who wake from anesthesia in severe pain, especially those who are not expecting pain, are already disadvantaged in terms of being able to control their pain—again, fear and anxiety intervene to make pain control difficult.

Every anesthetic plan should incorporate planning for postoperative analgesia. The first question is whether a nerve block, a spinal, an epidural, or a plexus catheter treatment would help. This decision must also involve whether these treatments can additionally be used intraoperatively to provide analgesia or even anesthesia. How does the risk weigh against the benefit of these procedures when considering their role for both intraoperative and postoperative use? When one of these techniques is chosen, it becomes necessary to plan the exact injection regime. Will a nerve block last into the postoperative period or need to be supplemented by systemic analgesics? Can spinal analgesia be prolonged by using an opioid in the injectate? How will an epidural or other catheter treatment be used so that both its intra- and postoperative effects are optimized?

Despite the usefulness of neural blockade and catheter treatments, the vast majority of cases are managed without these. Planning for intraoperative and postoperative analgesia remains important. The basic principles of opioid-sparing and preemptive analgesia, described later, are incorporated into the analgesia plan. Opioids are the only strong analgesics capable of controlling severe pain with systemic use, and they are the mainstay of both intra- and postoperative pain relief. Opioids can be used liberally during anesthesia but less liberally postoperatively because of side effects. Individual practices vary, but the goal of a pain-free emergence should always be incorporated into the plan. Provided they are not contraindicated, nonsteroidal anti-inflammatory drugs (NSAIDs), including injectable ketorolac or other adjuncts, can be used when predicted pain is mild or as an adjunct to opioids for moderate to severe pain.

Pain Assessment

Pain is a complex, multidimensional symptom resulting from a combination of tissue damage and nociception, previous pain experience, personal beliefs, culture, and mood. This explains why patients with the same degree of tissue damage can differ widely in their pain reports. Because there is no objective measure of pain, we must trust that patients' report of pain reflects their pain experience or distress. Acute pain is relatively straightforward to assess because, unlike chronic pain, it generally bears a predictable relationship to obvious tissue damage. The level of postoperative pain tends to change rapidly throughout the postoperative course, especially early after surgery; therefore, a policy of regular assessment of pain using simple measurement tools is the best way to ensure that pain treatment can be appropriately titrated. Side effects and recovery milestones are also measured to optimize analgesia and recovery.

Many institutions record pain as a "fifth vital sign" on the vital signs chart, especially since the introduction of the JCAHO mandate; others have a policy of recording a pain level in the medical record. Wherever it is documented, the recorded level is useful for communication between medical professionals and for assessing trends in an individual's pain level.

Verbal numeric rating scales are most commonly chosen to measure acute pain. The patient is asked to rate pain on a numeric scale, usually 0-10, where 0 is no pain and 10 is the worst pain imaginable. Verbal descriptive scales could also be used, but they are difficult for postoperative patients to cope with. Examples would be "mild, moderate, or severe" or "mild, discomforting, distressing, horrible, or excruciating." Visual analog scales are used in research, and unlike verbal numeric rating scales, they are validated for this purpose. The patient marks on a measured line labeled "no pain" at one end and "worst pain imaginable" at the other. Visual analog scales are rarely used in clinical practice, again because compliance is difficult for postoperative patients. Although the use of pain scales is necessary for documentation purposes, the clinical assessment of pain should always include asking patients to describe their pain and their pain experience since the last assessment. This is likely to reveal more information than the simple level. At the same time, the clinician should assess pain location, radiation, quality, and etiology. It is always important to expand the assessment to identify or exclude sources of pain not accounted for by the primary and expected source.

Assessment of pain in children, especially young children, is much more difficult. Although developmentally normal children older than 6 to 7 years of age will respond to adult assessment tools, younger children than this need a different approach. Infants, neonates, and very young children (0-4 years) are the most challenging. Children as young as 3 years old may be able to express pain and point to a painful area, but they are unlikely to be able to rate their pain. Children younger than this are unable to express pain other than by crying or screaming, posturing, grimacing, palmar sweating, and respiratory and heart rate changes—signs that are not specific to pain. Older children (4-7 years) are usually able to rate their pain, but it is helpful to use a scale designed for children, such as the Wong-Baker faces pain scale (Fig. 72-1). Parents are usually better able to recognize the presence of pain in their children than any of the other caregivers, and they should always be asked to contribute to the assessment process.

Basic Principles

Two important principles influence our choice of drug, dose, route of delivery, and timing for the treatment of acute pain. These are the principles of preemptive analgesia and opioid sparing.

Preemptive Analgesia

Preemptive analgesia is described by Igor Kissin as "an anti-nociceptive therapy that prevents establishment of altered processing of afferent input, which amplifies postoperative pain."21,22 In simpler terms, a preemptive analgesic intervention is one that stops or alters pain transmission so that pain will not become amplified by the nervous system. Kissin's definition embraces the fundamental principles necessary for understanding the concept of preemptive analgesia: (1) that the central nervous system is capable of changing so that pain becomes either improved and worsened via central processes such as desensitization and sensitization, and (2) that alterations in sensory and pain transmission can effect such changes. The idea that pain treated early in its course is easier to control than established pain is not new; in fact, it is a truism passed down through generations. The new idea is that in addition to psychologic factors, which remain important especially in humans, a central process—central sensitization—can be altered by altering afferent input using analgesic drugs or neural blockade. Animal studies showed strong and convincing effects after preemptive interventions including nerve blocks and systemic analgesics.23 Human studies have been less convincing.24,25 But the idea that our interventions before, during, and after surgery could attenuate pain responses is too attractive to lay to rest. Studies are still being conducted, and the clinical role of preemptive analgesia is still uncertain and much debated.

A systematic review published by Møiniche et al in 200224 found no support for preemptive analgesia when 80 trials assessing preemptive administration of epidural analgesia, systemic opioids, NSAIDs, local infiltration, and N-methyl-D-aspartate (NMDA) antagonists were grouped and analyzed, finding no overall benefit to any of the preemptive treatments. (The last intervention may have a special effect in that central sensitization is mediated by NMDA receptor activity).26,27 In all 80 trials selected here, preincision treatment was compared with postincision treatment. Despite this disappointing result, Ong et al went on to conduct a new meta-analysis, which included 10 newly published RCTs.25 Their methods differed from those of Møiniche et al in several respects: They included trials that compared the preemptive intervention with no intervention as well as the comparison with postincision interventions,28 they used a different approach for analyzing pain scores, and they used different exclusion criteria (fewer studies met the inclusion criteria). And in this case, several preemptive treatments—epidural analgesia, NSAIDs, and wound infiltration—were strongly effective. Preemptive opioid and NMDA antagonists were not effective. Note that Ong's analysis of NSAID preemptive effects needs reexamination after Reuben's studies were found to be fraudulent (see section Prevention of Chronic Postsurgical Pain).

The preemptive analgesia concept may seem simple: Stop central sensitization and pain will improve. But it is far from simple. Central sensitization itself is a complex process; NMDA antagonists may alter central sensitization but not necessarily because they are given prior to versus after an incision; opioids may actually cause central sensitization under some circumstances;29-31 neural blockade does not always succeed in preventing all afferent input from a surgical process.21,28 It is no surprise, then, that study results seem conflicting, and meta-analyses do not provide clear answers to the question of whether preemptive analgesia has an important clinical role. We have a lot to learn about how preemptive analgesia studies should be designed and how to combine them rationally in meta-analyses. Until we have better answers, we should be guided by preliminary data that suggest there is little to lose and possibly much to gain, from early use of neural blockade and NSAIDs, although the benefit of early NSAIDs may have as much to do with the pharmacology of NSAIDs (particularly delayed peak effects), as with any effect on central sensitization, the core mechanism proposed for true preemptive analgesia.32

Opioid-Sparing and Multimodal Analgesia

Opioid drugs mimic endogenous analgesic responses, so it is not surprising that they are our most powerful analgesics. Centuries of opioid (opium) use have provided humans with relief from pain and suffering. Yet, throughout this long history, there have been concerns about the opioids' toxic effects, particularly their addictive effects and the potentially fatal side effect of respiratory depression. The extent to which opioids have been used for the relief of pain has fluctuated, not only throughout history, but also between cultures. In the United States and other industrialized nations, opioid use for severe pain has been strongly encouraged, especially since the 1980s when it was recognized that antidrug regulations had severely inhibited opioid use to the detriment of pain control. Current teaching states that good pain relief during and after surgery, using opioids whenever necessary, can improve surgical recovery.33 Pain relief is an important component of accelerated recovery programs that aim to achieve early return of mobility, coughing, and bowel and bladder function, and to restore normal physiologic functioning as rapidly as possible thereby reducing complication rates.34-36 Yet herein lies a paradox: Opioids are our most effective analgesics; but they can delay recovery because of their sedative effects, their nauseating effects, and most importantly, by slowing bowel activity. Opioid-sparing interventions reduce opioid requirements, and in some cases they obviate the use of opioids altogether. Such interventions include the use of adjunctive medications, epidurals, and other nerve blocks, as well as nondrug interventions. The concept of "multimodal analgesia,"37,38 using several analgesic modes together, is similar to the concept of opioid sparing. It allows for administration of a combination of opioid and nonopioid analgesic drugs to be given that act at a variety of receptor sites within the central and peripheral nervous systems in an effort to minimize potent opioid use and therefore decrease opioid-induced adverse effects. The ability of nonopioid medications, such as gabapentin, to minimize the opioid analgesic requirements and associated side effects confers an opioid-sparing effect on these agents. Most published clinical data focus on the effects of adding a single therapeutic agent to an already established therapeutic analgesic regimen. More recently, data have begun to emerge using combination approaches to manage acute and postoperative pain such as local anesthetic infiltration of the surgical site, NSAIDs, NMDA antagonists, and α2-Δ-receptor modulators such as gabapentin and pregabalin. NSAIDs and selective cyclooxygenase (COX)-2 inhibitors consistently reduce opioid consumption postoperatively.39 Studies of NMDA antagonists have produced variable results, but there is evidence that low-dose ketamine has an important role as an adjunct to opioids in multimodal analgesia.40 Recent studies have shown an opioid-sparing effect by pregabalin in both hip arthroplasty and laparoscopic cholecystectomy patients, making it a useful adjunct in both same-day and short stay surgical procedures.41,42 Pregabalin has some advantages over gabapentin in that it has higher bioavailability and more linear pharmacokinetics that allow it to have more rapid onset. Local anesthetic infiltration into the surgical site has also shown to be clinically effective in a combination regimen for multimodal analgesia when treating acute postoperative pain.

Prevention of Chronic Postsurgical Pain

Apart from their role in reducing perioperative pain and analgesic requirements, a role in reducing chronic postsurgical (incisional) pain has been proposed for both preemptive and multimodal analgesic protocols.43,44 The hypothesis that postsurgical pain can be prevented or attenuated by perioperative analgesic interventions is appealing, yet the scientific evidence supporting the hypothesis remains inconclusive at present.45 Most convincing is the evidence that conduction blockade (paravertebral and epidural) is helpful for reducing persistent pain after breast surgery and thoracotomy. Evidence also suggests that surgical technique has an important role in minimizing persistent pain after surgery, not simply careful technique, but specific modifications such as use of minimally invasive and muscle- or nerve-sparing techniques and lightweight mesh (for hernia repairs).43,45

The present discussion would not be complete without mentioning the case of Reuben, a formerly respected and much published clinical investigator who fabricated data for 21 of his published studies in respected anesthesia journals. The studies all concerned perioperative analgesic interventions, and several supported the claim that adjunct and multimodal analgesia is capable of reducing postsurgical pain. The fraud is believed to be one of the largest known cases of academic misconduct, and the story was reported in the lay press as well as the scientific literature. The collateral damage was enormous because every paper, systematic review, and meta-analysis incorporating Reuben's studies had to be reexamined. Despite the enormity of this task, the greatest damage was, of course, to the credibility of the academic endeavor.

Although the damage the Reuben fraud has caused to the broader concept of multimodal analgesia is minimal, except in loss of credibility, the fraud has actually forced a reexamination of the concept that preemptive or multimodal approaches can reduce late pain. Countless studies and meta-analyses, many of which do not include Reuben's studies, support the opioid-sparing effect of epidural, regional anesthesia, NSAIDs, and other adjuncts. However with respect specifically to the concept that immediate or late pain can be prevented through the timely administration of nonopioid analgesics, Reuben's studies did have some influence so that reexamination became necessary. White et al, in an editorial in Anesthesia and Analgesia published in May 2009, examined each of the discredited studies and their potential influence.46 They concluded that the area where reexamination is necessary concerns the ability of NSAIDs (notably the selective COX-2 NSAIDs), and neuropathic pain medications (notably venlafaxine and pregabalin) in multimodal regimes to provide preemptive or long-term benefit after orthopedic surgery.

Systemic Treatments

Opioids

Opioids remain the mainstay of acute pain treatment, despite concerns about their side effects. Systemic opioids are used alone or in combination with other analgesics.47,48 Generally, higher doses of parenteral opioids are given during the first 24 to 48 hours after surgery when pain tends to be severe and patients are not able to tolerate oral medications. The intravenous (IV) route is the preferred parenteral route because most hospitalized patients already have IV catheters in place. Later, oral opioids, often in combination preparations such as Percocet (oxycodone with acetaminophen), replace the parenteral opioid. Side effects may limit opioid use, but there are few absolute contraindications, and true allergy to opioids is rare. Meperidine should not be used with monoamine oxidase inhibitors (MAOIs).

Choice of Opioid

Morphine is the main constituent of opium, it was the first opioid alkaloid to be identified, and it is the standard opioid to which other opioids are often compared. Codeine is the only other naturally occurring opioid in common use. The other familiar opioids are semisynthetic, derived from opium constituents (hydromorphone, hydrocodone, and oxycodone) or synthetic, synthesized de novo (meperidine, methadone, fentanyl, and fentanyl derivatives). These drugs are all opioid agonists, chiefly at the μ-opioid receptor. Other opioids are mixed agonists/antagonists or partial agonists (buprenorphine, butorphanol, pentazocine, nalbuphine) and are generally used for treating chronic rather than acute pain (or addiction), at least in the United States. They are not described in detail here. Naloxone is a pure opioid antagonist and can be useful to treat opioid overdose. The commonly used opioid-agonists are listed, with recommended doses, in Table 72-1. Opioid effects are summarized in Table 72-2. The opioids described next are μ-agonists with essentially similar effects. Choice of opioid depends as much on the familiarity and preference of the treating physician as on the generally subtle differences in pharmacology between these drugs.

Morphine is the standard, most widely used of the opioids, and in some countries, it is the only available opioid. It is the least lipophilic opioid, which delays its peak effect (occurs at 20 min after IV injection) but makes it a good choice for epidural administration when a widespread effect is desirable. Morphine is highly metabolized and has at least 2 active metabolites (morphine-6-glucuronide and morphine-3-glucuronide), which can delay normalization after morphine administration, especially during renal compromise and continuous administration. (Morphine metabolites may also contribute to morphine toxicity and morphine-induced hyperalgesia in the special circumstance of sustained use.) Morphine induces histamine release, and rapid bolus injection may produce local erythema, hypotension, or rarely, bronchospasm. The normal duration of a standard dose (10 mg) is 3 to 4 hours. Morphine has poor oral bioavailability, and the oral dose is 3 times the parenteral dose. Morphine may cause biliary and urinary tract spasm, and a different opioid is often substituted (meperidine or fentanyl) during treatment for gallstones and renal stones or during biliary tract and urinary tract surgery, although recent studies suggest that the effect is an opioid effect, not specifically a morphine effect.49,50

Codeine is less potent, although more constipating, than morphine. Constipation limits the recommended dose to 30 mg, which has only mild analgesic (and respiratory depressant) effects. Codeine is available for oral use only and is commonly found in combination analgesics such as Tylenol 3 (acetaminophen with codeine). In many countries, codeine is available over the counter because it is considered to have low abuse potential and a good safety record. Its main use is for pain of moderate severity, especially for children.

Hydromorphone (Dilaudid) is a useful alternative to morphine. For ill-defined reasons, many patients seem to prefer hydromorphone and claim they feel more clear headed, less dizzy, and less nauseated. At present, these can only be considered anecdotal observations. There are no active metabolites; therefore hydromorphone is a good choice for continuous administration, especially in patients with renal compromise.

Hydrocodone is used for mild to moderate pain, most commonly in the oral combination formulation Vicodin (hydrocodone with acetaminophen).

Oxycodone is not available for parenteral use in the United States, although it is elsewhere. It is a familiar opioid in the United States in the form of Percocet (oxycodone with acetaminophen) and widely used for the treatment of moderately severe acute pain, including home treatment. More recently, oxycodone was formulated as a long-acting preparation (OxyContin), which has become a useful alternative to morphine and MS Contin (long-acting morphine) for the treatment of cancer and selected chronic pain. OxyContin and other long-acting oxycodone preparations have also been used for the treatment of acute pain, to aid sleep at night and function during the day. OxyContin availability has been limited by constraints associated with its popularity as a drug of abuse.

Meperidine (Demerol) was widely used for sedation and analgesia during procedural treatments, particularly in the office setting, as well as for hospital treatment of intra- and postoperative pain, until in the early 1980s several problems with the drug became clear. The first was normeperidine toxicity. Normeperidine is a toxic metabolite capable of inducing central nervous system excitation and seizures, and liable to accumulate in the elderly and those with impaired renal function. The second was the propensity of meperidine to cause addiction, probably related to its lipophilicity and rapid onset of both analgesia and euphoria. Published guidelines began to state that meperidine should not be used routinely,7,8 and many hospitals withdrew it from their formularies. When it is available, meperidine remains a useful drug, and its ability to produce fast-onset analgesia and euphoria is its great advantage, especially during the treatment of postoperative pain. It also has an idiosyncratic and little understood advantage for treating postoperative shivering. It must, however, be used cautiously, being particularly mindful of normeperidine toxicity that can occur with repeated or continuous use. Meperidine has mild anticholinergic, antihistaminic, and local anesthetic effects. There may be a dangerous interaction (serotonin syndrome) with MAOIs producing seizures, coma, and possibly death, even after a single meperidine dose.

Methadone is a complex drug, rarely used for acute pain, except in patients already treated with this drug. Important considerations are its multiple interactions with other drugs, especially with antibiotics and antifungals, and its propensity to prolong the QTc interval, especially at high doses.51,52 Refer to more detailed texts if you are faced with a patient requiring high-dose methadone or receiving methadone as part of a complex drug treatment regime.

Fentanyl use in the postoperative setting is largely confined to PCA and epidurals, where its high lipophilicity and short duration can be used to advantage.

Tramadol (Ultram) is an interesting drug with weak opioid activity and additional norepinephrine and serotonin reuptake inhibition. It has low abuse potential and is not a controlled substance. It is widely used in Europe to treat acute and postoperative pain but is less favored in the United States, where it is only available for oral use. Its use for severe pain is limited by the fact that it has a ceiling effect, but it may be useful for mild to moderate pain, particularly in patients who refuse opioids or tolerate them badly.

Adverse Effects

Respiratory depression is the most feared of the opioid side effects, and rightly so, because this is a potentially fatal side effect. The possibility of respiratory depression produces a real conflict when trying to balance effective analgesia with safety. A key issue is to understand the likelihood of the event, and therefore the risk, and to provide adequate monitoring in high-risk situations. The immediate postoperative period is a period of high risk: The patient is often opioid naive, has been given multiple sedating drugs during surgery, may be weak, and may have a high analgesic requirement. Neonates and infants are always at high risk because of their immature nervous systems, propensity to apnea, and poor ability to metabolize opioid drugs. The elderly display similar risks. Patients established on a stable opioid regime are at lower risk. The level of monitoring required is a matter of judgment, whether this consists of frequent checks by a nurse or the application of a monitor such as an apnea monitor or pulse oximetry.

Other side effects are less catastrophic but can significantly compromise the success of opioid therapy. Patients may prefer to be in pain than feel disorientated, dizzy, or nauseated; physicians may undertreat pain rather than delay hospital discharge because of ileus or nausea. Secondary treatments such as antiemetics may help, but it is the principles of opioid sparing that play a key role in minimizing opioid side effects and optimizing pain control.

Tolerance, Dependence, and Addiction

It is important to understand tolerance, dependence, and addiction, and the differences between these 3 phenomena. Drug tolerance arises when an increase in dose is required to achieve the effect of an initial dose. Opioid tolerance arises through a combination of receptor desensitization (nonassociative tolerance) and psychologic factors (associative tolerance).53,54 Apparent tolerance could also result from opioid-induced hyperalgesia (OIH).31,55,56 Patients receiving chronic opioid therapy could manifest either or both pharmacologic tolerance and OIH, and in fact, the 2 may be difficult to distinguish clinically because both are manifest as increased opioid dose requirement, and both result from biologic adaptations to continued opioid use.57,58 The phenomenon of OIH becomes increasingly important during acute and postoperative pain management as more and more patients present for surgery with opioid refractoriness caused by the neuroadaptations to it.59 The management of these patients is described in the section Special Populations. The development of tolerance and hyperalgesia during acute treatment is rare but can occur, especially when opioid infusions are used (eg, remifentanil during surgery or opioid infusions on an intensive care unit).60-62 NMDA antagonists such as ketamine may have an increasingly important role in the treatment of acute and postoperative pain because they seem to be effective for attenuating hyperalgesia, OIH, and opioid tolerance.63-65

Dependence, also known as physical dependence, arises after chronic opioid use, is thought to reside in norepinephrine pathways of the locus ceruleus, and results in a typical withdrawal syndrome when an opioid-dependent patient is deprived of opioid. The typical opioid withdrawal syndrome comprises central neurologic arousal and sleeplessness, irritability, psychomotor agitation, diarrhea, rhinorrhea, and piloerection. In the management of acute pain, withdrawal may occur if habitual doses are not maintained during acute pain treatment. Patients who use opioids illicitly are not typically honest about their use, so physicians must be watchful for the signs of withdrawal when there is any suspicion of illicit use. Treatment consists of reestablishing previous opioid levels, before a gradual taper. Clonidine can be a useful adjunct and effectively treats many of the manifestations of opioid withdrawal.

Addiction, a behavioral syndrome with a neurobiologic basis, virtually never arises out of hospital treatment of pain with opioids. Patients who are worried about the addiction risk can be assured of this.66 However, addicted patients present for surgery and with acute trauma-related pain, and they require skillful pain management. The management of these patients is described in the section Special Populations.

Nonsteroidal Anti-Inflammatory Drugs and Acetaminophen

The NSAIDs are a group of drugs that inhibit cyclooxygenase (COX) (an enzyme in the arachidonic acid pathway), thereby inhibiting prostaglandin and thromboxane production (Fig. 72-2). Prostaglandin and thromboxane actions are summarized in Table 72-3. Inhibition of prostaglandin and thromboxane accounts for the analgesic and anti-inflammatory effects of NSAIDs, as well as for their adverse effects. Acetaminophen is not strictly an anti-inflammatory drug but is included here because it shares many of the properties of the NSAIDs. In contrast to the true NSAIDs, which are polarized and therefore do not readily cross the blood-brain barrier, acetaminophen is nonacidic and crosses the blood-brain barrier. Its action resides mainly in the central nervous system where prostaglandin inhibition produces analgesia and antipyresis. Its peripheral and anti-inflammatory effects are weak. Commonly used NSAIDs and their doses are listed in Table 72-4.

The NSAIDs and acetaminophen are commonly used over-the-counter analgesics, and have a long history of use for mild to moderate surgical pain, especially during home treatment. They are also used with opioids in oral combination formulations. The advent of ketorolac, the first injectable NSAID with Federal Drug Administration (FDA) approval for use in surgical patients, triggered a surge in interest in the use of NSAIDs as sole analgesics and as adjuncts for the treatment of moderate to severe acute and surgical pain. Ketorolac is a potent NSAID, unfortunately with a side-effect profile that reflects its potency, which can be used as a sole analgesic, even for severe pain. The NSAIDs and acetaminophen have emerged as useful adjuncts in multimodal analgesic regimes.

A new subclass of NSAIDs was recently released for clinical use: the selective COX-2 inhibitors. COX-2 is an inducible isoenzyme and a source of prostaglandins during inflammatory processes. COX-1, in contrast, is a constitutive isoenzyme and has protective effects on the stomach where it mediates the production of cytoprotective prostaglandins (Fig. 72-3).67 The selective COX-2 inhibitors were developed in the hope of being able to reduce NSAID side effects, particularly the damaging gastrointestinal (GI) effects. Early clinical trials and clinical experience confirmed the analgesic efficacy and favorable side-effect profile of these drugs with regard to their effect on the gastric mucosa and their platelet effects, although some of the early trial results have now been brought into question. The great hope for these drugs was that they would replace standard NSAIDs and reduce complications, particularly NSAID-induced GI bleeding, which is thought to account for up to 100,000 hospitalizations and 20,000 deaths per year in the United States. However, these hopes have been crushed by the steady emergence of evidence that deleterious cardiovascular and thrombotic effects preclude the use of these drugs in many patients.68,69 In fact, most are now withdrawn, and the only selective COX-2 inhibitor on the market in the United States at the time of writing is celecoxib (Celebrex).

Adverse Effects and Limitations on Perioperative Use

Adverse effects of NSAIDs in surgical patients are listed in Table 72-5. Contraindications arise out of these adverse effects, listed in Table 72-6. Acetaminophen is relatively safe and not associated with the adverse effects listed for standard NSAIDs. The COX-2 inhibitors are less likely to cause bleeding (platelet effects), particularly GI bleeding (unprotected GI mucosa), but they carry the same risk as standard NSAIDs with the other listed adverse effects and additional cardiovascular and thrombotic risks. These drugs therefore have similar contraindications to the standard NSAIDs.

Patients should be advised to stop taking NSAIDs before surgery, chiefly because of their platelet effects and their propensity to increase surgical bleeding. Aspirin, whose platelet effects are not reversible, should be stopped for up to 10 days before elective surgery. Other NSAIDs have rapidly reversible platelet effects, and 24-hour cessation is probably sufficient, although 2 to 3 days cessation is usual. Acetaminophen and COX-2 inhibitors can be continued because they do not have platelet effects.

A factor that makes perioperative NSAID use relatively safe is the fact that they are used for a short period, and most adverse effect are associated with prolonged use. This is true of platelet, GI, and renal effects. Thus even in patients chronically treated with NSAIDs who have stopped the treatment to minimize surgical bleeding, short-term perioperative use may be appropriate. This brings into question the issue of timing of NSAID administration, which is linked to the drugs' liabilities. There are theoretical advantages to giving NSAIDs early in the surgical course, even preoperatively. These include the drugs' pharmacokinetics—for example, the peak effect of ketorolac may occur as late as 4 hours after administration—and the consideration that the drugs may have preemptive effects. However, for major surgery, where there is a likelihood of bleeding and/or hypotension with deleterious effects on clotting and renal function, it is probably better to wait until the extent of any derangement is clear. Surgical considerations are also important when deciding whether to use an NSAID. These include possible postoperative bleeding, especially into closed cavities such as the knee joint, as well as retardation of bone remodeling, a consideration after bone fusion especially of the spine.

Use of NSAIDs and Acetaminophen for Postoperative and Acute Pain

For mild postoperative or acute pain, NSAIDs or acetaminophen can be used as sole analgesics. In addition, even though they are considered weak analgesics, they have an important role as adjuncts and in multimodal analgesic regimes. Their mechanism of action (prostaglandin inhibition) means that they are synergistic with many other analgesic interventions, particularly with opioid analgesia. Multiple studies and meta-analyses confirm an average 30% to 50% opioid-sparing effect of NSAIDs.68,70-74 Whether this reduction in opioid dose with NSAIDs translates into improved recovery and morbidity is less clear. The most recent meta-analysis of 22 RCTs by Marret et al affirmed a reduction in nausea (30% reduction) and sedation (29% reduction), but effects on urinary retention and respiratory complications were inconclusive. Overall, studies assessing the effects on adjunctive NSAID use on recovery have had mixed results.70,71 The availability of the injectable NSAID ketorolac has extended perioperative use to the many patients who cannot tolerate oral medications after surgery. Initially, the side effects of this drug seemed unacceptable, but this was an effect of an initial recommendation to use a 60 mg first dose with 30 mg repeat doses. Now that the recommended dose has been halved (30 mg first dose, 15 mg repeat doses) and a 5-day limit has been placed on ketorolac use, the early problems of catastrophic bleeding (GI, surgical, and joint) seem to be resolved. Injectable acetaminophen is available and widely used perioperatively in Europe, but it is not available in the United States.

Novel Adjuncts

Although the NSAIDs are the most widely used and useful systemic analgesic adjuncts in postoperative pain regimes, there has recently been interest in testing other possible adjuncts.

NMDA receptor antagonists are known to reduce central sensitization, hyperalgesia, and opioid tolerance.27,31,55 This makes them theoretically an attractive option for treating acute pain: By reducing central sensitization they might reduce postoperative pain and possibly reduce the likelihood of developing chronic pain; they could also reduce opioid requirements and opioid tolerance. Ketamine is the most widely tested of currently available NMDA receptor antagonists (ketamine, dextromethorphan, and amantadine), but use has been limited by this drug's side effects (psychomimetic effects, including nightmares and hallucinations). These side effects can be reduced by concomitant use of benzodiazepines and by dose restriction; efforts to study the possible usefulness of ketamine as an adjunct analgesic for postoperative pain have centered on dose-finding and regime modeling. A recent detailed study suggests that although ketamine has demonstrable antihyperalgesic effects (as shown by skin measurements around the surgical incision), this may not translate into a useful opioid-sparing effect.35,75 A 2005 systematic review by Elia and Tramer incorporating 53 trials (2839 patients) found a small difference (<1 cm on a 0-10 cm visual analog scale) in postoperative pain level, a significant difference in opioid use during the first 24 hours after surgery, and no difference in opioid side effects.76 The highest risk of hallucinations occurred in awake or sedated patients receiving ketamine without a benzodiazepine. The role of ketamine and other NMDA receptor antagonists remains uncertain at present, although ketamine is being used increasingly as evidence is accumulating that it may have a useful adjunctive role in reversing hyperalgesia, OIH and opioid tolerance.63-65

Neuropathic pain medications such as anticonvulsants and antidepressants, as their name suggests, have their chief pain indication for the treatment of chronic neuropathic pain. Gabapentin has long been a useful adjuvant medication for chronic neuropathic pain, and its use for perioperative and acute postoperative pain is increasing. A more recent addition to this class of agent, pregabalin, also has antiallodynic and antihyperalgesic properties useful for treating neuropathic pain, which appear to make them beneficial agents in acute postoperative pain.

Gabapentin and pregabalin are inhibitors of the α2-Δ-subunit of the high-voltage-activated calcium channel. Their use reduces both neurotransmitter release as well as postsynaptic excitability. Data increasingly suggest that both of these agents have a role to play in the attenuation of postoperative pain. A systematic review of randomized controlled trials (RCTs) by Ho et al in 2006 demonstrated that a single preoperative dose of 1200 mg of gabapentin was associated with significantly decreased cumulative opioid consumption at 24 hours after surgery. Even when gabapentin was administered at doses less than 1200 mg, pain intensity was lower at both 6 and 24 hours, and 24-hour cumulative opioid consumption was less, although these findings did not reach statistical significance. In all cases, gabapentin was associated with an increased risk of sedation with less opioid-induced adverse effects such as pruritus and postoperative nausea and vomiting (PONV).77 A systematic review by Tiippana et al in 2007 supports that gabapentinoids effectively decrease acute postoperative opioid consumption and opioid-induced adverse events following surgery.78

A more recent systematic review by Dauri et al in 2009 analyzed the evidence supporting gabapentin and pregabalin use for postoperative pain. When administered solely as a preemptive analgesic, gabapentin provided better postoperative analgesia and decreased opioid use postoperatively than placebo in 6 of 10 RCTs. However, gabapentin did not reduce PONV compared with placebo in 14 RCTs.79 Although studies support the role of gabapentin and pregabalin versus placebo in reducing pain and opioid consumption, comparisons with other postoperative regimens are lacking. More refined data delineating the optimal dose and duration of therapy, including when to begin, do not exist as yet.

Patient-Controlled Analgesia

Simple although it seems, PCA technology represents a huge advance in acute pain management. Computer-controlled pumps allow patients to control their own injections and to do this safely, and microchips have made controllable pumps easily portable. PCA satisfies the needs of patients to receive pain medication easily and quickly when needed. Nursing time spent obtaining, checking, documenting, drawing up, giving, and monitoring frequent doses is minimized. The hospital's need to comply with JCAHO's pain mandate is greatly aided. Most moderate to severe postoperative pain in hospitalized patients can be managed satisfactorily using routine IV opioid PCA for 24 to 48 hours after surgery, with or without adjuncts. A proviso is that pain in the immediate postoperative period is controlled by nurse bolus injections until it is under adequate control; patient-triggered boluses using standard settings may be inadequate for treating immediate postoperative pain, and early failed analgesia can prove difficult to overcome. Ideally, patients should be educated in the use of PCA before surgery.

The Inherent Safety of PCA

Because of the ease with which each PCA bolus dose can be given, PCA dosing regimes were devised using small frequent doses. For example, a standard regime for morphine is 1 mg every 6 minutes. A maximum hourly limit can be used, as can a background infusion if desired (the latter is particularly useful at night so that the patient can sleep). Part of the logic was to avoid the large swings between high peaks and low troughs associated with less frequent and larger doses (Fig. 72-4) but another was to improve safety. Small patient-controlled doses are inherently safe because a single dose is too small to produce overt sedation or respiratory depression, and an obtunded patient will stop pushing the button so there will be no further dosing beyond this early warning stage. The inherent safety of PCA also means there is less need for monitoring and no need to use the intramuscular route (with its slower absorption) for the sake of safety. Naturally, no method of delivery of opioids is completely safe, so a degree of vigilance is always required. The inherent safety of PCA is lost if persons other than the patient are permitted to push the button.

Benefits of PCA

Many studies have been conducted since PCA became popular in the 1980s to assess whether the use of PCA results in better analgesia, lower opioid requirements, fewer side effects, better surgical outcome, or superior patient satisfaction. Two meta-analyses of PCA versus "conventional analgesia" (intermittent large-dose opioid injection) have been published, the first in 199380 and the second in 2001.81 The second analysis added 17 new trials to the first, but its results were essentially similar to those of the first. There was slightly better analgesia associated with PCA use (difference 5.6 on a 0-100 scale, p = 0.006, debatably not a clinically important difference) and a large difference in patient satisfaction favoring PCA (42% improvement; p = 0.02). There was no difference in opioid usage, side effects, or surgical outcome. Thus the overriding benefit of PCA seems to be that patients like it. It has, in fact, become a standard of care in most US hospitals.

Cost of PCA

The issue of cost of PCA is frequently analyzed and debated.82-84 It has proven difficult to come up with a global assessment of PCA costs and the cost savings associated with reduced nursing involvement, improved safety, and other factors. This is because all related costs, including the initial outlay or rental of the PCA systems and their hardware, and including nursing costs, vary from location to location. Overall, it seems that the literal costs to a hospital are slightly higher for PCA versus "conventional analgesia," but imponderables such as increased safety and greater patient satisfaction are hard to translate into cost benefits.

Patient-Controlled Epidural Analgesia

PCA technology can also be used for epidural analgesia. As with IV PCA, the main advantage is that patients like the sense of control offered by the patient-triggered pump.

Novel PCA Technologies

Using iontophoretic transdermal and intranasal delivery systems, new methods of providing PCA are now in development.85,86 The most advanced of these is the fentanyl patient-controlled transdermal system (PCTS), which is a noninvasive, needle-free, credit card size, self-contained drug delivery system. A small imperceptible electric current drives ionized drug across the normally impenetrable stratum corneum (outer layer of skin). The adhesive system can be placed on any patch of hairless skin, and for convenience, the upper arm is usually chosen. The system is preprogrammed to deliver 40 μg fentanyl per dose and can deliver up to 6 doses per hour for 24 hours, or 80 doses, whichever comes first. Each dose is triggered by the patient pushing a button on the system twice (twice so it is not triggered accidentally). Cost is likely to play a large part in whether this type of system is adopted for hospital versus home use, or acute versus cancer/chronic pain.

Epidural Analgesia

The provision of epidural analgesia is probably the single most important contribution that anesthesiologists currently make to postoperative pain management for individual patients. There are 2 reasons for this: (1) after major surgical procedures, particularly abdominal and thoracic, epidural analgesia has proven analgesic superiority,87 an effective opioid-sparing effect, and probably a beneficial effect on surgical outcome,34,88,89 and (2) the ease with which the alternative (systemic opioid) can be provided using PCA has diminished the role of anesthesiologists in the provision of standard treatment. A perplexing thought, however, is how much we truly understand the risk versus benefit of epidural analgesia, especially in an era of potent thrombosis prophylaxis and seemingly increasing numbers of epidural hematomas. How do we weigh rare but catastrophic outcomes against common benefits of debatable value? This is one of the most important issues we face as we continue to teach and encourage the use of epidural analgesia.

Indications

A review of indications for intraoperative epidural use is outside the scope of this chapter, but there are also independent indications for postoperative epidural analgesia. These include patients having thoracic or abdominal surgery, patients having lower limb surgery in whom early mobilization is important, patients having lower body vascular procedures in whom a sympathetic block is desirable, and patients with compromised cardiac or pulmonary function.

Contraindications

Epidural placement is always contraindicated in patients who refuse this option. It is also contraindicated in patients with coagulopathy, concurrent or planned treatment with low-molecular-weight heparin or with potent antiplatelet agents, and in patients with bacteremia or local infection at the insertion site. The presence of spine pathology is a relative contraindication: The placement may be technically challenging and the treatment may fail if distorted anatomy prevents good distribution of the epidural medications. Neurologic disease is also considered a relative contraindication because if there is a change in neurologic status, which is not uncommon after the stress of surgery, diagnosis of the deterioration can be confused in the presence of an epidural.

Drug Choices and Drug Effects

Drugs injected or infused into the epidural space diffuse and spread according to their pharmacokinetics. The epidural space is complex because it contains arteries, veins, and lymphatics that are capable of absorbing drugs into the systemic circulation, nerve roots on which drugs can act directly, and fat in which drugs can form a reservoir (Fig. 72-5). The intrathecal space is a close neighbor of the epidural space, and diffusion into this space brings the epidural drugs across to more nerve roots, to the spinal cord, and, if spread is extensive enough, into the ventricular system of the brain (Fig. 72-6). When planning epidural infusion regimes, one has to consider not only local effects but also distant effects, as related to the individual drug's epidural pharmacokinetics.

Local anesthetics act directly on nerve axons to block sodium channels and thus block saltatory conduction. They can block nerves in the epidural space itself, or more widely once they cross into the intrathecal space. Because of the arrangement of nerve fibers within the nerve roots, with small fibers lying outside larger fibers, small fibers are the first to be blocked and the most sensitive to local anesthetics. This characteristic of the spinal nerve roots means that differential blockade can be achieved using low-dose local anesthetics to block only small fibers (C-fibers: sympathetic, pain, and temperature fibers) and high-dose local anesthetics to achieve total sensory blockade, often accompanied by motor blockade (α and β fibers). To achieve analgesia without sensory decrement, a low-dose local anesthetic (eg, 0.1% bupivacaine) is chosen for postoperative use. Epidurals with low-dose local anesthetic infusion are sometimes termed walking epidurals, a useful term because it emphasizes that patients can and should be walking around to hasten recovery and minimize complications.

Opioids have a completely different target: the opioid receptors in the dorsal horn of the spinal cord. Epidural opioid analgesia is also more likely to be complicated by systemic effects because, in contrast to the local anesthetics, which also undergo a degree of systemic absorption, there are clinically relevant systemic effects, especially when highly lipophilic opioids such as fentanyl are used. The degree to which opioids are absorbed systemically versus into cerebrospinal fluid (CSF) versus onto spinal cord receptors depends almost entirely on their lipophilicity (Table 72-7). As can be seen from Table 72-7, morphine is several times less lipophilic than all the other commonly used opioids. So it tends to be less readily absorbed into the systemic circulation and therefore provides a better selective spinal effect, and it also tends to spread more within the intrathecal space because it will favor staying in the watery medium of the CSF. This has advantages (more widespread spinal analgesia) and disadvantages (a higher risk that the drug will reach the ventricular system and cause respiratory depression; see Fig. 72-6). However, more lipophilic opioids such as fentanyl tend to have much greater systemic absorption (less selective spinal analgesia), and absorption onto spinal cord receptors tends to be localized, with poor spread. There is a great deal of experience using morphine and other opioids safely and effectively, and individual institutions must choose suitable regimes on the basis of the published experience and their ability to adequately follow and monitor patients receiving epidural analgesia.

The addition of clonidine to the local anesthetic and opioid has been found to significantly improve the quality and duration of neuraxial analgesia.90 The effect is mediated by descending modulatory pathways to the spinal dorsal horn. Despite the low doses used for neuraxial administration, systemic side effects (hypotension, bradycardia, and sedation) can occur. Dose-finding studies are still underway, and the therapeutic window for useful analgesia without side effects seems to be narrow. A reasonable regime uses a 1 to 2 μg/kg bolus followed by 0.4 μg/kg per hour.

Management Principles

The management of epidural catheters should always be under the direct supervision of anesthesiologists. Patients should be seen daily to ensure that catheters and medications are working effectively and possible complications are recognized, and recognized early. Pain reports should be satisfactory, and side effects such as pruritus, sedation, and changes in sensation or motor function should be carefully evaluated. Medication charts should be checked, especially for unintentional anticoagulant administration. Catheters and their insertion sites should be inspected for migration, integrity of the dressing, and for inflammation or back tenderness. Anesthesia personnel should make changes to the analgesic regime and administer specific medication as necessary. At the end of treatment, the anesthesia team should be responsible for pulling the catheter and ensuring it is removed intact. Nurses should be properly educated before they care for patients with epidural catheters. Important teaching points include typical medication doses and concentrations, anticoagulation issues, assessment parameters, the normal appearance of the catheter and catheter site, operation of the infusion pumps, common medication side effects that can be treated by them, and side effects requiring a call to the physician in charge.

If an epidural does not appear to be functioning well, it is first necessary to test whether the catheter is well positioned. The ideal way to do this is to attempt to produce a discernable level using a bolus dose of local anesthetic. However, the dose of local anesthetic needed to produce a sensory level may also induce hypotension, which would not be desirable in an unmonitored situation such as a regular floor, particularly if there are no means readily available to treat the hypotension. A surprisingly helpful test is to inject 5 to 7 mL of the analgesic infusion (low-dose local anesthetic), in which case no special monitoring is needed because the low-dose anesthetic is unlikely to produce extreme hypotension. If the catheter is well positioned, analgesia should be noticeably improved by the injection.

Once good catheter function is established, several approaches to improve analgesia can be taken. Bolus injections can be continued until good analgesia is achieved. The infusion rate can be titrated upward as tolerated. Epidural medication can also usefully be administered through a patient controllable pump, termed patient-controlled epidural analgesia. Systemic analgesic can be added to the regime. NSAIDs are useful adjuncts to epidural analgesia, especially when the epidural level does not cover the entire area of surgical pain, for example when the incision is high or when pain is referred outside the epidural area as occurs when chest tubes irritate the diaphragm and produce shoulder pain via the phrenic nerve. Systemic opioids (including PCA) can also be added, in which case it is preferable to remove the opioid from the epidural infusate, if used.

Managing Side Effects

Hypotension, mild sensory and/or motor changes, and urinary retention are the most common side effects of epidural local anesthetics, whereas pruritus, sedation, dizziness, and urinary retention are the most common side effects of epidural opioids. Most side effects are alleviated by either lowering the infusion rate or changing the drug or dose. Hypotension can become a considerable nuisance during epidural analgesia, even though the sympathectomy from low-dose local anesthetics is theoretically minimal. Some patients are particularly sensitive to the local anesthetic and manifest hypotension that does not respond to the primary measure of fluid replacement. For these patients, the local anesthetic can be removed from the epidural mix, or the epidural treatment may have to be abandoned altogether. Pruritus is a common side effect and usually responds well to antihistamine treatment. The mixed agonist/antagonist nalbuphine (Nubain) (5-10 mg IV, 4-6 mg hourly) also works well, as does low-dose naloxone infusion. Contrary to popular belief, nausea rarely occurs because opioid doses are low. Gut mobility is in fact improved by epidural therapy because of opioid sparing and the favorable effects of neuraxial local anesthetic on bowel motility. Urinary retention is common, and it is common practice to keep a Foley catheter in place until epidural therapy is discontinued.

Unilateral lower extremity numbness with occasional weakness or motor block occurs fairly frequently. It is often a result of the epidural catheter tip migrating along a nerve root so that the local anesthetic is concentrated in one area. Pulling the catheter back, or lowering the infusion rate, often rectifies the problem. However, one should always remain vigilant and continue to watch for more serious complications.

Complications

The most common complications are failed block/analgesia and postdural puncture headache (PDPH). Both are considered benign, although they may be devastating to patients who have committed to an epidural to optimize their surgical recovery. The exact incidence of these common complications is difficult to establish because reports vary, and the occurrences likely vary according to reporting and practice habits. Recent reports suggest that failed analgesia occurs in as many as 15% of cases.91-93 PDPH may occur in up to 86% of patients after accidental dural puncture (rate: 0.16%-1.3%).94 The incidence of other self-limiting neurologic complications such as radicular pain and peripheral nerve lesions is difficult to determine because these occurrences are rarely reported.

PDPH is thought to result from a small CSF leak secondary to accidental dural puncture. Typically, there is a delay in onset of the headache of up to 24 hours, so that the complication tends to manifest on the first postoperative day. Because PDPH tends to worsen on sitting up, and particularly on walking, it may not present itself until the patient gets out of bed for the first time after surgery. Other characteristics of the headache are that it tends to occur at the back of the head (occiput) and neck, and it produces a tight, pulling, and throbbing sensation. Conservative management consists of bed rest (up to bathroom only), plenty of fluids (IV or oral), and headache medication (NSAIDs, acetaminophen, Fioricet, caffeine, and theophylline all work well). If there is no resolution, or if conservative measures are contraindicated, a blood patch is recommended. This consists of an epidural injection of 20 mL of the patient's own blood (drawn under aseptic conditions). The exact mechanism by which an epidural blood patch works is uncertain but is probably either a pressure effect or a laying down of clot or fibrosis onto the puncture site.

Epidural hematoma with consequent paraplegia is extremely rare but a catastrophic complication of epidural therapy. Permanent paraplegia can occur even when a hematoma is diagnosed and treated in a timely manner, although early intervention usually reverses the neurologic injury. Persistent lower extremity neurologic changes or cauda equina syndrome (loss of bowel and bladder control) should always be taken seriously, especially if there is accompanying back pain and tenderness, which are cardinal signs of epidural hematoma and abscess although not necessarily present. Neurology can be consulted, but early imaging, preferably with magnetic resonance imaging (second choice computed tomography), should always be instituted. The presence of a space-occupying intracanal spinal lesion should always herald immediate transfer to the operating room for surgical decompression.

The incidence of epidural hematoma after neuraxial injection, catheter placement, or catheter removal has been estimated to be 1 in 190,000, with many of the reported cases associated with anticoagulant use.95-97 A rash of reports of epidural hematoma occurring after neuraxial interventions in patients receiving low-molecular-weight heparin alerted us to the dangers of epidural injections and catheters in patients receiving highly effective thrombosis prophylaxis. More problems followed when chronic treatment with potent and long-acting antiplatelet agents such as clopidogrel became more widespread.96 In view of the rapidity with which new agents are being introduced and the time lag before the extent of a problem can be assessed, we are left with a great deal of uncertainty about the safety of neuraxial procedures. Table 72-8 summarizes guidelines for managing epidurals in patients receiving anticoagulants. These recommendations are based on the consensus guidelines published by the American Society of Regional Anesthesia and Pain Medicine (ASRA). The guidelines are updated periodically and published on ASRA's Web site. Epidural bleeding is known to occur secondary to single-shot neuraxial techniques as well as neuraxial catheter insertion and removal, so recommendations are needed for the start and end of neuraxial therapy, as well as for starting anticoagulant therapy after neuraxial instrumentation or catheter removal.

Epidural abscess occurs less often than epidural hematoma, but it can be equally catastrophic and may cause permanent and serious neurologic injury, even death.98 Fever is a likely accompaniment to epidural abscess; otherwise it presents much as epidural hematoma. In fact, it may not be clear whether the mass is blood or pus until it is exposed during surgery. The mortality of spinal abscess can be as high as 18% (from case report).99 The incidence of epidural abscess secondary to neuraxial blockade is estimated at 1 in 250,000 in healthy patients, but 1 in 2000 in diabetic or immunocompromised patients.100,101 Other serious complications such as anterior spinal artery syndrome, transverse myelitis, and meningitis have been reported but are extremely rare.

Epidurals and Surgical Outcomes

In terms of truly understanding the benefit versus the risk of epidural analgesia in any individual patient, it is important to understand whether or not the benefit extends beyond that of the simple provision of excellent analgesia. Is mortality improved? Is recovery hastened? Are high-risk patients such as those with serious cardiac disease less susceptible to cardiac morbidity? Do the bowels recover quicker? Do epidurals reduce thromboembolism? Is postoperative pulmonary function improved? These are all questions that have been asked in countless trials often with conflicting results.74,89 Perhaps the clearest result is that epidural analgesia (particularly that from thoracic epidurals) does improve bowel mobility and reduce the period of ileus, with consequent earlier hospital discharge.102 There is also considerable support for improved pulmonary function.103-107 But support for a beneficial effect on other outcomes, particularly catastrophic outcomes, seems to be elusive. Large multicenter studies published relatively recently found no advantage to epidural analgesia in terms of mortality or major morbidity.105,107 Yet earlier trials and meta-analyses had suggested significant improvements in mortality and serious morbidity.103,104,106,108,109 Has some of the earlier advantage of neuraxial anesthesia and analgesia dissipated because they no longer provide the benefit of reduced thromboembolism in this era of improved thromboprophylaxis? And could general improvements in perioperative care such as better preoperative optimization, shorter acting anesthetic drugs, improved standards of vigilance and monitoring, and accelerated recovery programs have diminished the role of neuraxial anesthesia and analgesia?89

Epidural or No?

Despite all the blows against epidurals—incompatibility with modern thromboprophylaxis and dwindling returns in terms of improvements in mortality and major morbidity—they remain a beneficial and helpful intervention for selected patients undergoing invasive surgical procedures. For patients who accept the risks because of a promise of good postoperative pain relief, that promise is usually fulfilled. Patients with serious respiratory disease and those undergoing extensive lung resection benefit from opioid sparing and the favorable effects of epidural analgesia on pulmonary mechanics. Recovery can be hastened after bowel surgery because of favorable effects on bowel mobility. And patients with cardiac ischemia or failure can benefit from the moderate reductions in afterload and the negative inotropy and chronotropy of a well-managed thoracic epidural.

Nerve Blocks

The primary indication for nerve blocks is for the provision of surgical anesthesia. But there are secondary benefits in terms of pain relief in the postoperative period, and in some cases analgesic effects can be prolonged using plexus catheters. As nerve blocks wear off, there remains some degree of analgesia (probably from residual small fiber blockade), and surprisingly, these effects appear prolonged beyond the known half-life of the local anesthetic, especially in the case of distal peripheral blocks such as hand and ankle blocks. Certain nerves are accessible enough to be injected intermittently should the need arise: Femoral nerve blocks provide good but incomplete analgesia after knee surgery, and intercostal nerve blocks can be useful after thoracic surgery or chest trauma when epidural analgesia is contraindicated. The analgesia from single-shot neuraxial blocks can be prolonged by injecting an opioid at the time of placing the block. Injection or infiltration of local anesthetic into the wound by the surgeon can provide helpful analgesia during the early postoperative phase. These techniques are particularly useful for postoperative pain control: (1) prolongation of neural blockade using catheters, (2) neuraxial opioid injections, and (3) prolongation of neural blockade using catheters.

With the increasing number of procedures being performed in the ambulatory setting in the United States and elsewhere, it becomes especially important to ensure adequate postoperative analgesia. In fact, ambulatory surgical procedures now account for at least 50% to 70% of all surgical procedures performed throughout the United States.110 Regional anesthesia has several benefits in the ambulatory setting, and has grown in popularity among physicians and patients alike.36 The ability to provide an alternative to general anesthesia with associated rapid recovery, early analgesia, decreases in consumption of postoperative opioids, nausea, and sedation as well as improved mobility of the operative extremity make regional techniques a welcome and attractive addition to our armamentarium. A recent meta-analysis by Liu et al confirms that peripheral nerve blocks are associated with reduced pain, decreased postanesthesia care unit time, less postoperative nausea, and a decreased analgesic requirement.111 Reduction of patients' overall time in the ambulatory surgery setting was not demonstrated, but patient satisfaction was improved.

The type of regional technique chosen depends on the site and anticipated length of the surgery, any postoperative mobilization requirements, as well as a review of patients' comorbidities and possible contraindications to specific regional techniques (eg, peripheral techniques may be safer than neuraxial in patients with a coagulopathy).

Brachial plexus blocks are commonly used for upper extremity surgery. In general, these blocks are well tolerated and have few contraindications. Interscalene blocks produce ipsilateral phrenic nerve paresis for the duration of the blockade and should be avoided in patients who cannot tolerate a modest (25%) decrease in pulmonary function. The supraclavicular approach carries with it the risk of pneumothorax, and the infraclavicular approach may be useful particularly in the ambulatory setting. Regional techniques have transformed both total elbow and total shoulder arthroplasties from procedures commonly requiring several days of hospitalization into ambulatory procedures with the further option of continued analgesia via a perineural catheter.112,113

For the lower extremity, spinal anesthesia has largely been superseded by peripheral blocks in the ambulatory setting to minimize side effects such as urinary retention and inability to ambulate, all of which can delay discharge. Femoral, sciatic, and ankle blocks all provide excellent conditions for surgery as well as analgesia extended well into the postoperative period. As with the upper extremity blocks, plexus catheters can be placed at the time of block placement and used to prolong analgesia.

Intercostal nerve blocks can be useful when in the case of chest trauma; regional anesthesia helps improve pulmonary mechanics but epidural is contraindicated. Intercostal bupivacaine has been used successfully in patients with severe chest pain following rib fractures to provide sustained analgesia and a significant increase in both pulse oximetry saturation and peak expiratory flow rates.114,115

When a perineural catheter is placed, a PCA pump can be used to infuse via the catheter so that patients can control and adjust their own demand dosage of medication on the background of a preprogrammed continuous basal infusion. More recently, the development of new disposable pumps allows for patients to experience at home the type of analgesia that was recently available only as an inpatient. Recent data suggest that disposable pumps are as successful as battery-operated mechanical pumps, with fewer technical problems and higher patient satisfaction scores.116

Whatever peripheral nerve block is used to control postoperative pain, it is important to realize that a plan for either supplemental coverage of noncovered surgical areas or provision of breakthrough pain coverage is necessary and should be provided in addition.

Neuraxial Opioid Injections

Neuraxial morphine is safe provided dosing is reasonable and patients are appropriately monitored. A single shot of morphine into the epidural space (1-4 mg) or intrathecal space (0.1-0.4 mg) can provide prolonged analgesia (up to 24 h), but it carries a risk of delayed respiratory depression. Morphine is poorly lipophilic, tends to stay in CSF once there, and is subject to CSF bulk flow with passage to higher centers including the respiratory center (see Fig. 72-6). At the same time, the fact that morphine tends to remain in CSF is the reason that it produces excellent selective spinal analgesia (ie, good spread to spinal cord receptors). Single-shot neuraxial morphine is an excellent means of providing analgesia when there is no epidural catheter. Patients should be monitored in the same way as those receiving epidural opioid infusions. PCA can be used to provide supplementary analgesia if necessary, but for safety, only demand doses should be used, and continuous background infusions should be avoided.

Neonates, Infants, and Children

Throughout childhood, the management of pain presents special challenges. Very young children (neonates and infants) tolerate opioids poorly because they are prone to apnea, and their handling of this and other groups of drugs is altered by several factors, including slow conjugation by the liver. It is no longer acceptable to avoid giving opioids to neonates on the basis that they do not feel pain because there is now a great deal of evidence that they feel pain just as adults do and that untreated pain can produce an imprint that may have psychologic and even physical consequences later in life.117-120 Older children can also be scarred by traumatic and painful experiences, and it is sometimes difficult for adult caregivers to understand childhood and adolescent psychology and to communicate effectively so as to optimize pain management. Pain assessment is challenging, particularly in preverbal children (see the section Pain Assessment). Good pediatric pain management requires a great deal of time and patience and is therefore labor intensive; unlike adults, children, especially young children, cannot simply be left with a PCA pump. Unfortunately, the resources for providing excellent and safe pain care for children, which often must be done at the bedside, may be lacking other than in specialty hospitals and centers of excellence.

Planning for the Postoperative Experience

Anesthesiologists play a key role in preparing children and their parents for the postoperative period and the pain the children are likely to experience. First, they can explain what is going to happen and how pain will be managed and assessed. They can explain treatment options and help parents (and older children) select the right approach. Often an intervention such as an epidural, a single-shot caudal or a nerve block is helpful, and these options can be discussed. For children who are old enough for PCA (generally ≥6 years of age), this should certainly be discussed, and the children should understand how to use the pump. Parents should understand that the inherent safety of PCA will be bypassed if they press the button themselves and that it is better if they do not intervene other than to encourage the children to take doses. If the child has had past surgery, it is helpful to know what pain treatments worked well.

Treatment Choices

Essentially the treatment options available to adults are available to children. However, there are differences in the way neonates and infants metabolize drugs, particularly opioids, and these must be understood. In general, smaller, more frequent dosing is needed in the very young, but there is no substitute for referring to pediatric dosing guidelines. As in adults, opioids are the mainstay for treating severe pain, but given the safety issues, especially in young children, it is always preferable to reduce opioid requirements whenever possible by using adjunctive treatments such as NSAIDs and acetaminophen that are generally well tolerated. Epidurals, caudals, and other nerve blocks, even single shots, can be extremely helpful and are usually successful because the anatomy is straightforward. However, catheters can be difficult to maintain, especially if they are small-diameter catheters (when they have a tendency to block) and especially if the insertion site is within the diaper area (when the skin can become macerated and the catheter tends to slip out). Opioids can be given as a low-dose infusion in neonates and infants, which obviates the need for intermittent injections and is generally safe and effective, provided breakthrough pain is controlled by bedside bolus injections and the temptation to try and overcome uncontrolled pain by repeatedly increasing an infusion rate is resisted.

Safety

Maintaining safety while providing good pain relief is an especial challenge in children, chiefly because of their sensitivity to opioids. Young children require extra vigilance and great care with dosing. Dosing is calculated on a per kilogram basis, and because children's weights and sizes vary considerably, it always warrants careful calculation. Frequent checking of vital signs is mandatory in neonates and infants receiving opioids, and the use of respiratory rate and oxygen saturation monitors improves safety and is a standard of care for this population in many institutions.

The Elderly and Infirm

Just as children present a special challenge, so do the elderly and infirm. Again, this is a group of patients that is particularly sensitive to opioids. They are prone not only to opioids' respiratory depressant effects but also to their central dysphoric and euphoric effects. Opioid-induced confusion is common in this age group. The elderly are liable to become confused with other psychotropic drugs such as benzodiazepines, and they are sensitive to NSAID side effects, particularly GI bleeding and renal dysfunction. They can also develop a confusional state simply because they are moved from familiar surroundings. These factors combined make good pain control hard to achieve. Pain control may become a low priority because it is hard to assess pain in confused patients, and drugs are blamed for the confusional state and therefore withdrawn.

The elderly tend to appear stoical, and it is not clear whether they have a different threshold for pain, whether past experience has altered their attitude toward pain, or whether they truly do not feel pain to the same extent as younger adults. It is probably best to treat pain according to the patients' demands and not to project pain on them because of a preconceived notion that their pain must be worse than they say it is. PCA is not a suitable choice for confused patients, and it may not work well even in nonconfused elderly patients because they tend to undermedicate. A simple approach such as nurse-determined intermittent low-dose opioid with acetaminophen as an adjunct is often the best approach. Epidurals and nerve blocks are useful for reducing reliance on poorly tolerated systemic medications.

Opioid-Tolerant Patients

Pain specialists and acute pain services are increasingly preoccupied treating patients who are opioid tolerant and who get into difficulty when an episode of acute pain, notably due to surgery or trauma, proves refractory to standard treatments.56,58,121 Nonspecialists are often uncomfortable prescribing the high doses of opioid that may be needed to overcome acute pain in opioid-tolerant patients. There are several reasons that the numbers of patients presenting with opioid tolerance is increasing, some related to societal factors, others to changes in medical practice. (1) All efforts to control drug and heroin abuse by means of regulation have failed, and illicit drug trading, distribution, and use steadily increase as high illicit drugs prices and profits are maintained. (2) Prescription drug abuse increased exponentially during the 5 years before 2001,122 largely because opioid use for chronic pain was liberalized during the late 1900s, resulting in a greater availability of prescription opioids in homes, pharmacies, and on the streets. Addicts choose prescription drugs over illicit drugs because of their greater reliability and safety. (3) In the 1980s and 1990s, pain advocates, carrying the success of their much needed lobbying for the use of opioids for acute and cancer pain, extended the principle to the treatment of chronic noncancer pain. Thus today there are many more patients with chronic pain conditions treated with opioids than there were just 20 years ago. (4) Cancer is now a curable disease for many patients, and others experience long remissions. They often need treatment for pain associated with their primary disease, or with its treatments, and they may need this treatment for years. Because of a tradition of opioid use for cancer pain, opioids are being used for long-term treatment of cancer-related pain.

Managing pain in opioid-treated patients, whether they are substance abusers or chronic pain patients, can be extremely challenging.56,59,121 There are issues related to opioid tolerance and pain refractoriness, and a withdrawal syndrome could further complicate the clinical picture. In addition, there may be difficult behavioral issues in both groups of patients, so that the dedication and professionalism, particularly of the nursing staff, can be severely tested. The possibility that drugs are being diverted complicates matters even further. Prejudices set in about the veracity of pain complaints, there is a breakdown in trust, and the simplicity of treating pain according to the patient's report is lost.

Substance Abusers

Substance abusers are notorious deceivers, but they recognize that they put themselves at risk if they deny use and are usually honest about whether or not they are currently using drugs, including opioids. They are less honest about past use and unreliable in terms of reporting doses. The treating physician must be cognizant that opioid requirements for pain are unknown and can only be learned by titration; that a withdrawal syndrome may arise if doses are underestimated; and that manifestations of other abused drugs (cocaine, amphetamines, alcohol, etc) or withdrawal from these may also require treatment. Treatment for withdrawal syndromes is generally supportive, but in the surgical setting, opioid withdrawal should be treated with opioid. An α2-agonist such as clonidine may help reduce norepinephrine effects.

The period of hospitalization for surgery or trauma is not the time for tackling an entrenched addiction problem. Nevertheless, if the patient agrees (and often the person does not), the involvement of a psychiatrist or addictionologist during the hospitalization can be helpful. The surgical and pain teams should focus on optimizing pain management and direct the patient to the appropriate resource after discharge, assuming he or she wants treatment. An opioid-abusing patient may have an inadequate response to the opioids given for pain, and it is helpful to make maximum use of alternatives such as epidurals, nerve blocks, and NSAIDs. At the same time, opioids should not be withheld, particularly if there has been recent use.

Chronic Opioid-Treated Pain

Opioid tolerant chronic and cancer pain patients are among the most difficult patients to treat for pain after surgery, particularly if they are being treated with high doses of opioids and have become hyperalgesic or refractory.55,123 Despite the fact that there is now convincing evidence that the analgesic efficacy of high-dose opioid therapy may diminish over time and has unacceptable side effects such as gonadal suppression, there are still patients admitted to hospital on doses exceeding the common upper limit of 180 mg morphine or morphine equivalent per day.31 The addition of chronic intrathecal opioid treatment may further complicate and compromise the treatment of acute pain.

Principles of treatment are similar to those for substance abusers. The normal regime should be continued if possible, and if not, it should be converted to an IV regime. Nonopioid treatments should be used to as great an extent as possible, to minimize reliance on opioids. Neuraxial interventions should be avoided in patients with implanted stimulators or pumps, unless the position of the implanted system is clear, because the needle can cut or damage electrodes, leads, or catheters. PCA is again a useful mode of delivery for opioids because a background infusion can be used to whatever level is needed, and the ability to get extra bolus doses reassures the patients. High-dose boluses are likely to be needed.

The period of hospitalization is not a good time to change the chronic pain regime, and most efforts should focus on treating the acute pain. However, if the surgical or pain team is having difficulty managing a patient, it may be very helpful to talk to the patient's pain physician or whoever normally manages their pain. Sometimes issues are revealed that are relevant to the management of pain in the hospital but not necessarily clear from the patient or the patient's chart. It is also helpful to have a plan for treating pain after discharge, which could involve the patient's normal physician or could involve rehabilitation or a pain program.

Behavioral problems can arise in opioid-treated chronic pain patients, although these will be different from those in substance abusers. Chronic pain patients tend to have difficulty coping, and when they are faced with uncontrolled and excruciating pain, which they have not necessarily predicted, they can be angry, mistrustful, demanding, or uncooperative. They may display typical opioid-seeking behaviors (frequent demands for pain medication, refusal of all but opioid treatments, accusations against the medical staff for undertreating pain), and these behaviors appear much like addictive behaviors. Yet it will be difficult to determine whether the behaviors have arisen because of undertreated pain (pseudoaddiction) or because there is, in fact, an underlying addiction problem.

Pain Management in the Intensive Care Unit

Several challenges are associated with managing postoperative pain in patients requiring care in the ICU. In many cases, these patients are unable to communicate because of severe illness, extensive surgical trauma, or because they are ventilated and sedated. Continuous infusion is the most frequently chosen means of delivery for opioids, but prolonged use of opioid infusions may lead to rapid development of tolerance or opioid-induced hyperalgesia, making pain control and weaning from opioids difficult.30 One of the most obvious manifestations of opioid-induced hyperalgesia is the ICU patient treated with high-dose opioid infusion who cannot tolerate touch or even the sheets. Patients with compromised renal function may accumulate opioid metabolites, notably the active metabolite morphine-6-glucuronide and the toxic metabolite normeperidine. Meperidine is not a good choice for prolonged infusions; if morphine is used, the possible accumulation of active metabolites must be considered in the weaning protocol. Epidural analgesia is an extremely useful option that is opioid sparing and can help during weaning from ventilation.106,124-126 However, epidurals are often contraindicated in ICU patients because of a need to treat with anticoagulant therapy. Use of activated protein C, our current best agent for severe sepsis, also precludes the use of epidurals.

It is important to treat pain in ICU patients to reduce the anxiety associated with pain and the inability to communicate pain. When it is impossible to assess pain, as in heavily sedated or unconscious patients, it is reasonable to assess analgesic requirements on the basis of the amount of surgical or other trauma the patient has undergone. Ventilated patients can be treated with higher than normal doses of opioids (if desired) because there is no risk of respiratory depression. However, it is important to remember that prolonged infusions may lose their efficacy if tolerance or hyperalgesia develop, and all opioid-sparing strategies are helpful. Methadone may also be useful for prolonged ICU stays because there is less development of tolerance. It has recently been found beneficial to use the α2-agonist dexmedetomidine for ICU sedation, not only because of its hypnotic effects, but also because of its analgesic synergy with opioids (opioid sparing), and possibly a contribution to minimizing opioid withdrawal at the time of weaning.127,128 Dexmedetomidine (often in combination with remifentanil or other opioids) is gaining increasing popularity also as a sedating agent during airway and other relatively noninvasive procedures.129,130

Advances in surgery, anesthesia, drugs, and technology have combined to markedly reduce the trauma and morbidity of surgery. Part of this success can be attributed to improvements in pain management, both during and after surgery. In the 1980s, several changes revolutionized postoperative pain management. Portable PCA pumps became available, and gradually the use of PCA was adopted widely. Clinical use of neuraxial opioids began after endogenous opioids and opioid receptors were identified in the 1970s. It was soon recognized that neuraxially administered opioids could provide excellent targeted analgesia at low doses without the need for larger systemic doses and associated side effects, or the sensory and motor blockade from the high-dose local anesthetics previously needed for adequate epidural pain relief when local anesthetics were used alone. So-called walking epidurals began to be widely used for postoperative pain control. Several professional and political bodies, recognizing that good pain control was an important component of perioperative care, developed guidelines to propagate good evidence-based practice (Agency for Healthcare Policy and Research). Insurers, spearheaded by Medicare, also recognized the importance of good pain control and began to reimburse pain management services, allowing for the development of acute pain services. Later, pain advocates succeeded in enlisting JCAHO in the war on pain, and their 2001 mandate requires all accredited hospitals to have a systematic means of recognizing and assessing pain, and treating it appropriately. Postsurgical pain may not be fully conquered, and there is certainly room for more research and further improvement. But tremendous progress has been made since the not so distant days when uncontrolled pain meant that patients routinely remained motionless in bed for days after surgery, prone to complications of immobility such as thromboembolism, atelectasis, and protein wasting.

• Angst MS, Clark JD. Opioid-induced hyperalgesia: a qualitative systematic review. Anesthesiology. 2006;104(3):570-587.
• Ballantyne JC. Does epidural analgesia improve surgical outcome? Br J Anaesth. 2004;92(1):4-6.
• Brennan TJ, Kehlet H. Preventive analgesia to reduce wound hyperalgesia and persistent postsurgical pain. Not an easy path. Anesthesiology. 2005;103:681-683.
• Kehlet H. Postoperative opioid sparing to hasten recovery: what are the issues? Anesthesiology. 2005;102(6):1083-1085.
• Kissin I. Preemptive analgesia at the crossroad. Anesth Analg. 2005;100(3):754-756.
• Perkins FM, Ballantyne JC. Post-surgical pain syndromes. In: Stannard C, Kalso E, Ballantyne JC, eds. Evidence-based Chronic Pain Management. Oxford, UK: Blackwell; 2010.
• Practice guidelines for acute pain management in the perioperative setting: an updated report by the American Society of Anesthesiologists Task Force on Acute Pain Management. Anesthesiology. 2004;100(6):1573-1581.
• White PF, Kehlet H, Liu S. Perioperative analgesia: what do we still know? Anesth Analg. 2009;108(5):1364-1367.