Postoperative pain management in the United States was fairly standardized before the 1980s. Mild-to-moderate pain was treated with acetaminophen or a nonsteroidal antiinflammatory drug (NSAID), and moderate-to-severe pain was treated with intermittent, intramuscular opioids, as needed. In the 1980s, the options for managing postoperative pain expanded. When endogenous opioids and their receptors were identified in the late 1970s, the value of neuraxially administered opioids was realized, and epidural analgesia became popular. During the 1980s, infusion pump technology improved dramatically and intravenous patient-controlled analgesia (PCA) became a viable option for the treatment of acute pain. More recently, interest in regional anesthesia for the management of postoperative pain has increased, especially with the use of ultrasound guidance for the performance of these procedures.
Ready and colleagues introduced the concept of a formalized acute pain service to provide coordinated, interdisciplinary acute pain care. This prompted widespread use of integrated, interdisciplinary teams to provide inpatient acute pain services, as well as increased efforts to improve the process of patient care, thus improving pain-related patient outcomes. The importance of multimodal analgesia was recognized, and as a result, there was increased interest in the impact of nonopioid analgesics on acute pain control. This opened the door to further investigations into the use of nonopioid analgesics, including NSAIDs, acetaminophen, and gabapentin, in the perioperative period. The goal of therapy was not necessarily to replace opioids with these medications but rather to improve acute pain control, lower the incidence and severity of treatment-related adverse effects, and perhaps lower the incidence and severity of chronic pain after selected surgical procedures.
When the era of evidence-based medicine dawned at the end of the twentieth century, most existing trials of postoperative pain management compared newer approaches (e.g., epidural analgesia, PCA, and adjunctive use of NSAIDs) with conventional analgesia, often defined as systemic opioids. Therefore early attempts at formulating an evidence-based approach to postoperative pain management focused on assessing whether these new approaches offered superior analgesia or a better effect on surgical outcome compared with conventional analgesia. As a result, first attempts at providing evidence-based recommendations focused on assessing the outcomes associated with the use of a single approach to postoperative pain management.
There are many factors, some of which are not obvious, that can affect the results observed in clinical trials. The decisions made by health care providers throughout the surgical period can affect outcomes in a number of ways, from patient selection for the surgical procedure, to the performance of anesthesia and surgery, to the care provided in the immediate postoperative period. Therefore variability exists within and outside the context of pain care, and this variability can have an impact on the outcomes of pain-related clinical trials. This variability may also make it difficult to document the outcomes related only to the pain therapy of interest.
Many experts advocate the use of perioperative rehabilitation, often defined as the use of an integrated, interdisciplinary process of care for patients undergoing the same kind of surgical procedure. With perioperative rehabilitation, a team approach is advocated. Pain management is integrated into this care and is often multimodal, no longer relying only on a single pain therapy technique. Therefore the manner in which pain therapy is applied may well be another treatment variable to consider when the outcomes associated with pain therapy are evaluated.
It is hard to imagine that 20 years ago patients undergoing major surgery would stay in the hospital for up to 2 weeks (sometimes longer); would stay immobile in bed, unable to move because of severe pain; and would not be given water or food until the bowel function was regained. Not surprisingly, the incidence of thrombosis, embolus, infarction, and infection was much higher than seen today. Improvements in pain care have contributed to the recent improvement in surgical outcomes.
Postoperative care has evolved from passively waiting for recovery to actively encouraging a return to normal function. Although pain control is critical to rapid recovery, the clinician must balance the risk of harm associated with pain therapy with the potential for benefit. Although uncontrolled pain can, indeed, delay a return to normal function by making patients afraid to move, breathe deeply, or cough, systemic opioids may delay recovery by causing adverse gastrointestinal effects, increased sedation, and (rarely) compromise of ventilation.
An important role of postoperative pain management is to maximize pain relief while minimizing adverse side effects. Treatment options should be evaluated not only according to their ability to provide satisfactory pain control but also by their ability to promote recovery and rehabilitation. The choice of postoperative analgesia should be procedure-specific because analgesic efficacy is contingent on the specific type of surgery. In this chapter, we consider the evidence supporting the use of epidural analgesia, intravenous PCA, NSAIDs, and continuous peripheral nerve blocks (CPNBs).
Epidural analgesia can be accomplished by infusing a variety of medications (typically a combination of low-dose local anesthetics and opioids) into the epidural space. Epidural analgesia must be distinguished from epidural anesthesia , which implies dense epidural local anesthetic blockade that can be used as the primary anesthetic for surgery. Conceptually, the provision of epidural analgesia is an attractive means of minimizing the opioid requirement while providing excellent analgesia (especially with movement), thereby promoting recovery after surgery. Epidural opioid doses are much smaller than those required systemically (in the order of one tenth), and low-dose epidural local anesthetics, apart from producing analgesia without overt sensory/motor blockade or opioid-associated adverse effects, can have additional beneficial effects on bowel mobility. Does the evidence support the superior analgesic efficacy of epidural analgesia and its ability to promote recovery after surgery?
It is important to note that patient outcomes may vary based on how and when epidural analgesia is administered. When used during surgery, the benefit of epidural anesthesia likely relates to the impact on outcomes of profound neural blockade (e.g., lower incidence of thromboembolic events, lower incidence of graft failure in the case of major vascular surgery, less blood loss, lowering of the metabolic stress response, and a lower incidence of chronic pain). The timing of the administration of epidural analgesia relative to surgery (i.e., starting analgesia before, during, or after completion of surgery) may also affect patient outcomes, but the impact of timing on outcomes has not been clearly documented. This chapter concentrates on the benefits likely to pertain specifically to postoperative epidural analgesia.
Many of the early trials of epidural analgesia (during the 1970s and early 1980s) were small randomized studies that attempted to confirm the clinically apparent superior analgesia of postoperative epidurals compared with systemic opioid analgesia, and some of these studies assessed the impact of epidural analgesia on postoperative recovery. These early trials (and meta-analyses) overwhelmingly supported the superior analgesic efficacy of epidural analgesia compared with systemic opioid administration. Assessment of postoperative recovery focused on a variety of outcome measures including pulmonary function, bowel function, and patient mobility.
A goal of epidural analgesia is to restore normal physiologic function as rapidly as possible so that adverse outcomes presumed to be associated with prolonged immobilization and hospital stay can be avoided. As evidenced by small randomized controlled trials (RCTs) and subsequent meta-analysis (and, in some cases, confirmed by the large RCTs), epidural analgesia fulfills this goal extremely well. Epidural analgesia has been shown to promote early mobilization and reduce rehabilitation time, particularly after joint surgery. In addition, it has been shown to reduce pulmonary morbidity, reduce time to extubation after major thoracic and vascular procedures, reduce cardiac ischemia and dysrhythmia in high-risk patients, and reduce postoperative ileus, thereby reducing length of the hospital stay. The overall benefit of epidural analgesia on patients undergoing cardiac surgery remains controversial. A meta-analysis by Beattie and colleagues found a reduction in the incidence of myocardial infarction (MI) associated with the use of postoperative epidural anesthesia (odds ratio [OR], 0.56; confidence interval [CI], 0.30 to 1.03). However, another recent meta-analysis by Svircevic and colleagues demonstrated a statistically significant reduction in supraventricular arrhythmias and respiratory complications but failed to show any benefit in MI, stroke, or mortality.
Several clinical trials have been conducted to evaluate the impact of epidural analgesia on mortality rate and major morbidity (including major cardiac morbidity, pulmonary embolus, and stroke). Early results suggested that combined epidural and general anesthesia followed by postoperative epidural analgesia had a favorable effect on major morbidity and possibly also on mortality rate. The findings of Yeager and colleagues were particularly striking because they showed remarkable decreases in surgical morbidity and mortality rates attributed to epidural analgesia in high-risk patients undergoing major surgical procedures. Interestingly, this study was stopped after the completion of 53 patients by the monitoring committee because the committee believed that the observed outcomes favored the epidural treatment so strongly that it would be unethical to continue the trial. This study certainly contributed to the belief among many pain physicians that epidural analgesia improves surgical outcomes, particularly in sick patients.
It is important to note, however, that the validity of the results of the Yeager study has been called into question. Indeed, two large RCTs have now been published that failed to confirm the positive outcomes associated with the use of epidural analgesia reported earlier by Yeager. Both studies provided strong evidence regarding the efficacy of epidural analgesia. However, although both studies were designed and appear to have been powered adequately to confirm the results reported by Yeager, neither study demonstrated a reduction in major morbidity and mortality rates in high-risk patients after the use of epidural analgesia.
Epidural analgesia may play an important role after abdominal surgery. In this setting, epidural analgesia has been reported to lower the incidence of MI, stroke, and death in patients undergoing abdominal aortic surgery. A meta-analysis evaluated the impact of epidural analgesia versus systemic opioids after abdominal aortic surgery. This analysis included 13 studies involving 1224 patients. The epidural analgesia group showed significantly improved pain control on movement up to the third postoperative day. In addition, the postoperative duration of tracheal intubation and mechanical ventilation was significantly shorter by about 20%. The overall incidence of cardiovascular complications, MI, acute respiratory failure, gastrointestinal complications, and renal insufficiency were all significantly lower in the epidural analgesia group, especially in trials that used thoracic epidural analgesia. The incidence of mortality, however, was not reduced.
Another study provided indirect evidence that epidural analgesia may lower mortality rates after major surgery. This study examined a cohort of 3501 patients who underwent lung resection. These patients represented a 5% random sample of patients who underwent lung resection procedures between 1997 and 2001 and who were listed in the Medicare claims database. Multivariate regression analysis showed that the presence of epidural analgesia was associated with a significantly lower odds of death at 7 days (OR, 0.39; 95% CI, 0.19 to 0.80; p = 0.001) and 30 days (OR, 0.53; 95% CI, 0.35 to 0.78; p = 0.002). Interestingly, this study reported no difference in major morbidity rates.
It is easy to forget that the evolution of epidural analgesia has occurred alongside the evolution of postoperative management in general and that differences in serious morbidity and mortality rates that might be expected to emanate from the benefits outlined earlier may not be obvious because of improvements in postoperative care in general. A policy of early oral fluid administration, early nasogastric tube removal, and forced early mobilization, in combination with optimized pain management, has resulted in earlier hospital discharge and a decrease in the postoperative mortality rate when compared with that 20 years ago. It may be impossible to show the benefit of postoperative epidural analgesia in isolation; instead, studies of this mode of analgesia used with regard to its specific effects on certain outcomes (e.g., postoperative ileus); with attention to appropriate level of catheter placement, drug choice, and drug dose to achieve the desired outcome; and its combination with other aspects of postoperative care are needed before we can discount the value of epidural analgesia in terms of major morbidity and mortality rates. At the same time, major morbidity and mortality rates have become so low that very large numbers of patients will have to be enrolled to be able to detect a difference in morbidity between study groups.
In summary, the superior analgesic efficacy of epidural analgesia compared with conventional analgesia seems very clear, and benefits in terms of morbidity and length of hospital stay (by contributing to an accelerated return to normal physiologic function) have been demonstrated. Although epidural analgesia appears to have a positive impact on the incidence of a number of major complications related to surgery, it remains unclear whether epidural analgesia has a role in reducing perioperative mortality.
The use of sophisticated infusion pumps to enable patients to self-deliver doses of analgesic medications (PCA) was popularized in the 1980s when microprocessors became small enough to be incorporated into portable pumps. A majority of patients and nurses prefer PCA to nurse-administered “as needed” opioid administration: patients prefer it because of the greater control they achieve over their analgesic dosing, and nurses prefer it because of the convenience of this mode of analgesia. PCA is now available in most hospitals in the United States, and it has become an important tool to aid hospitals’ compliance with mandated pain assessment and treatment standards. PCA is most often used for the systemic delivery of opioids (intravenous PCA), but this method of drug delivery has also been used to deliver opioids, local anesthetics, and other drugs (often in combination) via other routes, such as via an epidural catheter or a regional nerve block catheter.
Intravenous PCA differs from conventional analgesia in two important ways: (1) provided the technique is used appropriately, peaks and troughs in serum analgesic level are less extreme, and analgesic administration is better matched to analgesic need; and (2) patients report a greater sense of control over their pain care. The questions to be asked are, do these factors result in improved pain control, lower opioid requirements, superior patient satisfaction with treatment, fewer side effects, and better surgical outcome? Intravenous PCA is compared here with conventional analgesia; the use of patient-controlled epidural analgesia (PCEA) in the management of postoperative pain is also increasing in popularity, but this use will not be addressed.
Three meta-analyses of PCA versus conventional analgesia have been published, one in 1993, the second in 2001, and the third in 2006. Apart from updating the first analysis, the second incorporated trials in which control group opioids were given by the subcutaneous and intravenous as well as the intramuscular route. Fifteen trials (787 patients) were included in the first analysis, 32 (2072 patients) in the second, and 55 (3861 patients) in the third. All but one trial (which used meperidine) in the first analysis used morphine in both experimental and control group patients (699 patients). In the second and third analyses, morphine was used in the majority of studies, but other opioids were also used, including hydromorphone, meperidine, piritramide, nalbuphine, and tramadol.
The first meta-analysis demonstrated that patients prefer PCA to conventional analgesia and that PCA had slightly better analgesic efficacy. The mean difference in satisfaction was 42% ( p = 0.02), whereas the mean difference in pain score on a scale of 0 to 100 was 5.6 ( p = 0.006). However, there was no difference in opioid use, side effects, or length of hospital stay.
Despite the passing of almost 10 years and the addition of 12 trials (1000 patients) to the first meta-analysis, the results of the second analysis differ very little from those of the first. Patients’ preference for PCA was confirmed, as was slightly better analgesic efficacy. In three morphine trials and one meperidine trial, PCA was preferred (relative risk [RR], 1.41; CI, 1.1 to 1.80). Combined data on pain intensity and relief from one piritramide, one nalbuphine, and eight morphine trials also demonstrated a preference for PCA (RR, 1.22; CI, 1.00 to 1.50). There was no difference in opioid use or side effects and no convincing evidence of a difference in surgical outcome, although the limited data (152 patients) available on pulmonary function did suggest an improvement.
The third meta-analysis included yet more studies (55) and patients (2023 receiving PCA and 1838 receiving conventional analgesia). Even with an increase in the number of trials and patients, the results were again similar, in that PCA was demonstrated to provide better pain control and patient satisfaction than conventional analgesia. However, patients using PCA consumed higher amounts of opioids than the control subjects and had a higher incidence of pruritus but had a similar incidence of other adverse effects. There was no difference in the length of hospital stay.
Another meta-analysis evaluated PCA compared with conventional analgesia after cardiac surgery. This study used patient-reported pain intensity as the primary outcome and cumulative opioid use, intensive care unit and hospital length of stay, postoperative nausea and vomiting, sedation, respiratory depression, and all-cause mortality rate as secondary outcome measures. The authors identified 10 RCTs involving 666 patients. Compared with conventional analgesia, PCA significantly reduced the visual analog scale at 48 hours but not at 24 hours after surgery. PCA increased cumulative 24- and 48-hour opioid consumption. Ventilation times, length of intensive care unit stay, length of hospital stay, patient satisfaction scores, sedation scores, and incidence of postoperative nausea and vomiting, respiratory depression, and death were not significantly different.
Do these meta-analyses represent the best evidence about the utility of intravenous PCA compared with conventional analgesia? Certainly, the meta-analyses help by providing a quantitative summary of existing data. However, because many of the trials contributing to these meta-analyses were small, treatment effects may have been distorted because of deficiencies inherent in small trials, including type I error, distortions that can possibly be compounded in meta-analyses. Another problem encountered here (and, indeed, in many epidural trials) is that neither patients nor assessors were blinded to treatment; thus there is a high likelihood of assessor bias, which might be expected to exaggerate treatment effects. One should also be concerned with the degree of differences observed in the analysis. Some meta-analyses have demonstrated a small improvement in analgesic efficacy and possibly pulmonary function with PCA use. However, the clinical significance of these findings needs to be questioned not only because of the small effect size but also because of the weakness in the design of the contributing studies as a result of blinding.
It is also worth noting that in real life there are wide variations in factors such as patient education and nursing workload that have a profound effect on the doses of analgesic actually received and thus on the efficacy of either method. In addition, there may be significant differences in outcomes in patients participating in clinical trials evaluating pain outcomes when compared with patients receiving routine postoperative care outside the context of an analgesic trial. Thus the efficacy of PCA compared with conventional analgesia is likely to differ among trials and real life, individual patients, and institutions.
Patients’ preference for PCA seems to be an important reason that PCA has been established as the standard of care for routine management of moderate-to-severe postoperative pain. In view of the lack of evidence of any other real advantage to PCA, other than a slight improvement in analgesic efficacy, it seems that the reason that patients prefer PCA is that it provides them with a sense of autonomy and control over their own analgesic management. In today’s health care climate, patient preference is an important and valid reason for a treatment choice.
Given the lack of evidence of other benefits, one has to ask whether the cost of PCA is justified. Preliminary cost–benefit analyses suggest that postoperative analgesia with PCA is more expensive than conventional analgesia, despite the hope that nursing involvement would be reduced and reduced nursing costs would offset the increase in equipment costs. However, the results of cost-effectiveness studies are often based on cost data specific to one institution at a specific time; therefore they may not be valid at other institutions or at different times.
Nonsteroidal Antiinflammatory Drugs
NSAIDs have been demonstrated to be effective analgesics for the treatment of pain after surgery. This has been proven in single-dose studies in mild-to-moderate pain, as well as in multiple-dose studies in moderate-to-severe pain. NSAIDs have clearly been demonstrated to have an opioid-sparing effect.
When the use of NSAIDs are considered in the postoperative period, several issues are key. First, does the addition of an NSAID improve pain control, lower the incidence of opioid-induced adverse side effects, or both? Second, does the addition of an NSAID present new risk of harm to the patient? Third, are there any benefits to the use of cyclooxygenase-2 (COX-2) selective agent in this setting?
After major surgery, NSAIDs alone cannot provide effective pain relief. Therefore they are added to other pain therapy, such as systemic opioids. When given in combination with other opioids after surgery, NSAIDs result in better pain relief and lower opioid consumption. A 2005 meta-analysis evaluated the administration of NSAIDs on morphine PCA. This analysis included 33 trials with 1644 patients. In the trials evaluating multiple-dose regimens of NSAIDs, the average reduction in 24-hour morphine consumption was 19.7 mg, which was equal to a 40% opioid-sparing effect. In addition, the use of an NSAID lowered pain intensity from approximately 3 to 2 on the 10-cm visual analog scale when compared with morphine PCA alone.
The addition of an NSAID with the resultant reduction in opioid consumption may not lower the overall incidence of adverse events. It seems clear that the incidence and degree of respiratory depression is reduced, but improvements in pulmonary function (less opioid-induced hypercapnic responses) have not been convincingly demonstrated. The adjunctive use of NSAIDs reduces the incidence of nausea in several studies, although an equal number of studies do not show any benefit. The literature is equivocal about whether opioid sparing by NSAIDs promotes rapid recovery. A limited number of studies demonstrate accelerated recovery in association with less nausea and sedation, improved mobility, and earlier return of bowel function, but others fail to show any benefit in terms of recovery.
A meta-analysis evaluated the effect of NSAID administration on PCA morphine side effects. This study included 22 randomized, double-blind clinical trials published between 1991 and 2003, with 1316 patients receiving NSAIDs and 991 patients receiving PCA morphine only. This study demonstrated that NSAIDs significantly decreased the incidence of postoperative nausea and vomiting by 30%, and the incidence of sedation by 29%. Pruritus, urinary retention, and respiratory depression were not significantly decreased by NSAIDs.
NSAID use may be associated with a number of potential adverse events, including inhibition of platelet function, alteration in renal function, peptic ulceration, and alterations in bone healing. However, it appears that short-term use of NSAIDs around the time of surgery may not be associated with a compromise of bone healing. The risk of NSAID-induced adverse events is higher with higher doses and longer durations of therapy. In addition, the risk of harm is higher in the elderly.
A meta-analysis evaluated the effects of NSAIDs on postoperative renal function in adults with normal renal function. This analysis included 23 trials with 1459 patients. Perioperative administration of NSAIDs reduced creatinine clearance by 16 mL/min (95% CI, 5 to 28) and potassium output by 38 mmol/day (95% CI, 19 to 56) on the first day after surgery compared with placebo. However, no significant difference was seen in serum creatinine level on the first day (0 mmol/L; 95% CI, −3 to 4). No significant reduction in urine volume during the early postoperative period was found, and no cases of postoperative renal failure required dialysis. Other studies have demonstrated that the risk of adverse renal effects is increased in patients with pre-existing compromise of renal function, hypovolemia, hypotension, or the concomitant use of other nephrotoxic drugs.
Concern has developed over the last several years about the cardiovascular consequences of NSAID administration. This concern was triggered by evidence that the COX-2 inhibitors may lack the thrombotic-protective and cardioprotective effects of aspirin and other standard NSAIDs but now extends to the demonstrated deleterious effects of NSAIDs in general on cardiac function and blood pressure, especially in susceptible patients. A meta-analysis that included 55 trials with 99,087 patients evaluated the impact of COX-2 selective agents on the risk of MI. The overall pooled OR for MI risk for any coxib compared with placebo was 1.46 (95% CI, 1.02 to 2.09). This study concluded that celecoxib, rofecoxib, etoricoxib, valdecoxib, and lumiracoxib were all associated with higher MI risk compared with placebo. The pooled OR for any coxib compared with other NSAIDs was 1.45 (95% CI, 1.09 to 1.93). Another meta-analysis reported that all NSAIDs increase the risk of MI and cerebrovascular accidents, and COX-2 selective agents confer the highest risk.
In summary, perioperative NSAID administration is associated with significant opioid-sparing effects and a resultant reduction in several opioid-induced side effects. Other than the differences in the effect on platelet function, there appears to be little advantage to the use of COX-2 selective agents, and the use of these agents may be associated with an increased risk of cardiovascular adverse events.
Continuous Peripheral Nerve Block
Interest in CPNBs in the perioperative setting has increased within the past few decades. Similar to epidural analgesia, peripheral nerve block catheters are an attractive means of minimizing opioid requirements while providing excellent analgesia, thereby promoting increased mobility and rapid recovery after surgery. First described in 1946, the peripheral nerve catheter has evolved from a needle inserted through a cork attached to a patient’s chest to percutaneous insertion of a catheter directly adjacent to a peripheral nerve. The peripheral nerve catheter serves to continuously provide analgesia (typically local anesthetics) to the affected region postoperatively. Single-injection peripheral nerve blocks are also useful in the postoperative setting, but they are limited in their duration, usually lasting less than 24 hours. As with PCA administration, a discussion of CPNBs must involve the following questions: do these regional interventions result in better analgesia, lower opioid requirements, increased patient satisfaction, fewer side effects, and better surgical outcome?
Previous RCTs have shown that the addition of a CPNB greatly decreases postoperative pain, as well as opioid requirements. Unfortunately, these studies have included a relatively small sample size and have failed to show statistical significance in the reduction of pain in all time periods.
A meta-analysis evaluating the efficacy of CPNBs in the reduction of postoperative pain, undesired side effects (including nausea, vomiting, sedation, and motor/sensory blockade), opioid use and patient satisfaction reviewed 19 RCTs involving 603 patients. CPNB provided superior postoperative pain control when compared with systemic opioids ( p < 0.001), as well as a lower incidence of side effects (i.e., nausea, vomiting, and sedation). Total opioid consumption was also significantly less ( p < 0.001) with the peripheral catheter (20.8 mg morphine; 95% CI, 18.5 to 23.1; n = 165 patients) compared with opioid analgesia (54.1 mg morphine; 95% CI, 50.8 to 57.4; n = 174 patients). This meta-analysis demonstrated that, compared with opioid analgesia, CPNB provides superior postoperative analgesia, decreased opioid requirement, and decreased side effects such as nausea and vomiting. Further investigation is needed to determine the role of CPNBs in the postoperative setting, including questions regarding optimal catheter placement, technique, and equipment used in placement (i.e., stimulating-catheter versus ultrasound), as well as the different types of infusions and their relationship to toxicity and adverse effects.