Chapter 16: Analgesia, Sedation, and Neuromuscular Blockade

In addition to providing vital organ support and treatment of the underlying condition responsible for admission to the intensive care unit (ICU), the health care team must assure patient comfort and limitation of further stress on the patient. Failure in this regard may result in physiologic derangement and unwanted cognitive side effects. In order to attain these goals, a thorough knowledge of the pharmacokinetics (what the body does to the drug) and pharmacodynamics (what the drug does to the body) of a variety of medications is necessary; many of these agents are used exclusively in the ICU or by anesthesiologists in the operating room.

Over the past decade, an expanding body of basic science and clinical research has changed the way intensivists think about the treatment of pain and the management of sedation in the ICU. The use of newer drugs, such as dexmedetomidine as well as newfound indications for older drugs (eg, ketamine and lidocaine), have improved our ability to provide appropriate analgesia and sedation in the ICU. Recent trial results have provided high-quality evidence for use of neuromuscular blockade, despite its side-effect profile, in specific clinical circumstances (eg, severe acute respiratory distress syndrome [ARDS]). Older paradigms of analgesia and sedation in the ICU have evolved to incorporate patient-centered outcomes, such as quality of life and functional status after ICU discharge in survivors. This has been accomplished by aggressive assessment and treatment of patient discomfort and focusing on maintaining the ability of the patient to interact with caregivers and their environment through lighter sedation. The goals of this chapter are to underscore the necessity of analgesia and sedation in the ICU; describe pharmacologic properties of important analgesics, sedatives, and neuromuscular blocking drugs (NMBDs); and describe clinical strategies that improve patient outcomes.

Epidemiology of Pain in the ICU

Pain is common in the ICU, but is more difficult to assess than in acute care patients because of a common inability for critically ill patients to subjectively express pain. The inherent subjective nature of pain is well described by the International Association for the Study of Pain, which defines pain as an “unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.”¹ However, the inability to communicate pain does not mean it is not being experienced by the patient. In fact, it has been reported that the majority of critically ill patients (both medical and surgical) do experience pain^1,2—which patients describe as a significant source of stress, if not the most significant source of stress during their ICU stay, even when asked 6 months later.³

In addition to an unpleasant experience, undertreated pain may have negative somatic and long-term consequences. The pain response increases plasma catecholamine levels, causing vasoconstriction, impaired tissue oxygenation, and increased myocardial oxygen demand. Pain can cause neuroendocrine activation resulting in a catabolism state, breakdown of muscle tissue, and impaired immune function, which can increase susceptibility to infection. In addition to acute anxiety and fear, acute pain is associated with chronic pain, posttraumatic stress disorder, and lower health-adjusted quality of life after ICU discharge.

Physiology and Anatomy of Pain

The pleiotropic effects of pain are due to the complicated processing of pain from stimulus to cerebral cortex. There are four described elements of pain processing:

• 1) Transduction. Noxious stimuli are converted to an action potential.

• 2) Transmission. Action potentials are conducted via afferent neurons.

• 3) Modulation. Afferent pain signals are altered by efferent neural inhibition via neurotransmitters, especially in the dorsal horn of the spinal cord or by augmentation via neuronal plasticity (eg, central sensitization).

• 4) Perception. Integration of afferent pain signals in the cerebral cortex.

Specific analgesic drugs and drug classes, which act at each of these four pathways, are presented in Table 16–1. Multimodal analgesic strategies are those considered to target more than one pathway simultaneously. Pain chronicization—a transition from acute pain to chronic pain—is a feared consequence of painful critical illness, and much research has addressed neuronal plasticity in this pathophysiology. Fortunately, therapeutic options have been developed to decrease the risk of this transition, and will be discussed further as follows (see Nonopioid Analgesics).

Strategies for Pain Management

Assessment of Pain

Techniques for appropriate pain management must be individualized to each patient, starting with an appropriate assessment of its severity. This assessment should be more detailed than a simple visual analog scale (VAS). If the patient is able to adequately communicate pain, the following should be assessed: site, onset and timing, quality, severity, exacerbating and relieving factors, response to analgesics, and assessment of pain with movement, breathing, and cough.

As mentioned earlier, critically ill patients may not be able to self-report, so behavioral assessments should be used. Based upon their psychometric properties (ie, reliability and validity), the Critical Care Pain Observation Tool and Behavioral Pain Scale are currently recommended for use in adults over other reported scales. It should also be noted that the use of vital signs as the sole surrogate for pain is strongly discouraged. Although more validation testing in certain patient populations who may fail to exhibit typical pain behaviors (eg, neurologically injured) is warranted, implementation of behavioral scales is associated with lower resource consumption—such as days of mechanical ventilation—and ICU length of stay (LOS).³

Opioids—Systemic opioids are traditionally the cornerstone of postoperative and critical care pain management. As shown in Table 16–1, opioids' sites of action affect 3 of 4 pain-processing pathways. Opioids act as ligands at G protein-coupled opioid receptors, namely, the μ (mu), δ (delta), and κ (kappa) receptors. These receptors are located peripherally, in the spinal cord dorsal horn as well as at various locations in the brain. Opioid receptors are located on primary afferent neurons and inhibit release of nociceptive substances, decrease neurotransmitter release in the spinal cord, and activate descending inhibitory neurons. The μ-receptor is thought to be responsible for much of the peripheral analgesia caused by opioid agonists. There are a number of opioid receptor subtypes that have been discovered and are expressed on a variety of different cells, some unrelated to the central nervous system (CNS) and peripheral nervous system (such as on leukocytes and vascular tissue). This broad expression of opioid receptors is likely because there are number of endogenous opioids. The complicated interaction and broad expression of opioid receptors has led to opioids being implicated in such diverse pathophysiology as increased rates of cancer recurrence and modulation of inflammation. Intensivists should, therefore, have an appreciation not only for the acute adverse effects of opioids (respiratory depression and sedation), but also more subtle, long-term effects of opioid therapy. For more information regarding specific opioids, see Table 16–2a and 16–2b.

The ideal method of opioid administration will vary considerably with the clinical context and include opioids by mouth or parenteral administration either via intermittent intravenous or via patient-controlled infusion pumps (PCA, patient-controlled analgesia). For patients requiring mechanical ventilation, bolus or continuous infusion of opioids can treat both pain and ventilator-associated anxiety.

PCA—For awake, alert patients with moderate to severe pain—most commonly postoperative patients—PCA is an appropriate choice. Before initiating a PCA order, the patient must be able to understand how to appropriately use the PCA apparatus. To determine the PCA prescription, the following parameters must be specified: (1) opioid, (2) incremental (or demand) dose, (3) lockout interval, (4) background infusion rate (if any), (5) 1- and 4-hour dose limits, and (6) bolus dose administration (for breakthrough pain). A reasonable starting prescription in an opioid-naïve patient would be morphine with an incremental dose of 1 to 2 mg, a lockout of 6 to 10 min, no continuous infusion rate, a 4-hour maximum dose of 30 mg, and a bolus dose of 2 to 4 mg every 5 min for 5 doses.

Although PCA has gained significant popularity for its ease of use, when compared to intermittent bolus opioids, outcome studies show mixed results. There is some evidence that patient satisfaction is improved, possibly because of the patients' increased sense of control over pain.⁴ In general, there seems to be little change in analgesic efficacy, with patients reporting similar VAS pain scores with PCA and intermittent IV bolus analgesia. For patients that are taking chronic opioids when treated with a PCA, total dose of opioid was thrice higher than patients not taking opioids previously.⁵ Interestingly, patients with opioid tolerance had increased rates of sedation compared with opioid-naïve patients, which suggests the therapeutic index for opioid-tolerant patients may be narrower than opioid-naïve patients.

Nonopioid Analgesics

Acetaminophen and Nonsteroidal Anti-inflammatory Drugs

Acetaminophen is a ubiquitous centrally acting cyclooxygenase inhibitor with a long history of use outside the ICU, but very little research has been applied to critically ill patients. Much of the research in non-ICU patients showed modest (but statistically significant) reductions in mild and moderate pain. The notable exception regards research studying the IV formulation of acetaminophen, which became available in the United States in early 2011. Pharmacokinetic studies showed increased plasma and cerebrospinal fluid levels of acetaminophen after IV administration compared with oral or rectal administration. Although studies results are mixed, decreased VAS pain scores, opioid side effects, and early extubation following major surgery have all been demonstrated with IV acetaminophen.⁶ Given the common scenario of treating patients in the ICU with moderate to severe pain and concern for impaired absorption from the gastrointestinal tract or when side effects of opioids are particularly harmful, IV acetaminophen may play an important role, but attention should be paid to the increased cost of the IV formulation.

Many acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs) are used in acute postoperative pain and pain associated with critical illness; these drugs work primarily though cyclooxygenase and prostaglandin synthesis inhibition. All are generally effective in decreasing moderate and severe pain in equal-analgesic doses, but their utility is limited by their side effect profiles. NSAIDs can indirectly cause constriction of the afferent renal arterioles and inhibit platelet function (both in clinically significant ways), so their use is relatively contraindicated in patients at risk for renal injury and significant bleeding.

Ketamine

Ketamine, historically used as an IV dissociative general anesthetic with hypnotic and analgesic properties, has been the subject of much research in the treatment of acute pain. Ketamine acts primarily though N-methyl-d-aspartate (NMDA) receptor antagonism, which plays an important role in pain modulation and prevention of the transition of acute pain to chronic pain. Prevention of pain chronicization is a particularly useful characteristic of ketamine, especially in those patients that are opioid tolerant, as evidenced by a randomized controlled trial (RCT) of patients undergoing spine surgery that showed reductions in pain scores at 48 hours and 6 weeks after ketamine versus placebo infusion intraoperatively.⁷ Ketamine also may reduce opioid-related side effects, including nausea and vomiting. Because analgesic adjunct doses of ketamine (0.05-0.4 mg/kg/h) are significantly smaller than anesthetic doses (1-3 mg/kg/h), psychomimetic reactions from ketamine are significantly less common at these doses, and only slightly more common than placebo. Other side effects of ketamine (sympathetic stimulation and salivation) are not usually observed with analgesic doses.

Lidocaine

Lidocaine is a well-known sodium channel antagonist local anesthetic that also has antiarrhythmic, general anesthetic, and analgesic (both antinociceptive and antineuropathic) properties. It has been studied extensively in the perioperative setting as an IV infusion, with mainly mixed results. Cytokine levels decrease after lidocaine infusion, suggesting that inflammation causing nociceptive pain may be decreased. Lidocaine has also been shown to modestly decrease VAS scores, opioid requirements, and duration of postoperative ileus postoperatively.⁶ A recent study enrolling patients with complex spine surgery showed a significant improvement in subjective physical health months after surgery with lidocaine infusion versus placebo (a secondary outcome measure).⁸ Lidocaine has significant cardiac toxicity and his cleared by the liver, and therefore, should be used with caution in critically ill patients with hepatic dysfunction.

Gabapentin and Pregabalin

As with lidocaine and ketamine, the calcium channel blockers gabapentin and pregabalin have been useful in treating neuropathic pain. There is considerable research demonstrating increased analgesic efficacy of gabapentin as part of a regimen that also includes opioids in patients with chronic pain. There is also evidence for effectiveness in reducing opioid requirements and opioid-related side effects in patients with acute postoperative pain. Data supporting use of gabapentin and pregabalin in the ICU is limited, but a study evaluating patients with pain related to Guillain-Barre demonstrated decreased pain scores and opioid consumption in the ICU compared with placebo. In a meta-analysis of postoperative patients (not necessarily in the ICU) the number-needed-to-treat-to-benefit (NNT) to achieve 50% reduction in pain for gabapentin (NNT 11) is higher than that of naproxen (2.7), ibuprofen (2.5-2.7), or oral oxycodone 15 mg (4.6).^1,9,10,11,12 The most common side effect of gabapentin is sedation, but this effect is relatively rare—number needed to harm was 35.^7,13 Taken together, these data suggest gabapentin and pregabalin are safe and effective adjuncts in the treatment of pain in the ICU.

A summary of nonopioid analgesics is provided in Table 16–3.

Regional Analgesia and Nonpharmacologic Techniques

Regional analgesia—targeted administration of local anesthetics and analgesics to certain anatomic locations—may have considerable advantage in the management in acute postsurgical and traumatic pain. Lower doses of medication can be delivered in a targeted fashion to decrease the risk of side effects of systemic delivery of medication, and analgesia is often significantly improved compared to IV opioids. In certain ICU scenarios regional analgesia should be considered: (1) thoracic epidural analgesia in open abdominal aortic aneurysm surgery and (2) thoracic epidural analgesia in traumatic rib fractures, especially in the elderly. Using thoracic epidural analgesia in both these scenarios is supported by clinical practice guidelines from the Society of Critical Care Medicine,³ and have improved clinical outcomes—mainly by decreasing respiratory complications.

Although data supporting regional analgesia as a means to reduce morbidity is limited, in a broader array of scenarios regional analgesia decreases VAS pain scores, decreases systemic opioid requirements, decreases opioid side effects (postoperative ileus and urinary retention), and it may improve patient satisfaction.⁶ In a recent meta-analysis, regional anesthesia decreased the incidence of persistent postoperative pain 6 months postsurgery after thoracotomy and breast surgery by two-thirds (odds ratio = 0.33, 95% confidence interval [CI]: 0.2-0.66).¹⁴

In addition to pharmacologic analgesic therapy, there are a variety of nonpharmacological analgesic techniques that may be helpful to patients. These techniques include music therapy, relaxation, therapy, family presence, distraction, massage, and deep breathing. Although empiric evidence of their efficacy is mixed and precludes high-level recommendations, their low cost and safety make them a valuable consideration within the context of a multimodal analgesia strategy in the ICU.

Sedation Pharmacology

Once determining that sedation is necessary for a critically ill patient, the selection of sedative should be based on the predicted length of required sedation, the reason sedation is necessary (eg, agitation, increased intracranial pressure, or alcohol withdrawal) as well as the pharmacodynamics and pharmacokinetics of the drug (Table 16–4). This strategy should involve patient factors, such as hepatic and renal dysfunction, hemodynamic stability, and inferred changes in drug volume of distribution (V_D).

Critical illness can significantly alter sedative pharmacokinetics in a variety of ways. Most sedatives undergo hepatic metabolism, many with a high hepatic extraction ratio, and therefore, decrease in hepatic blood flow that occur commonly in patients with shock, after abdominal surgery, or with increased intra-abdominal pressure can significant decrease drug clearance. Sedatives are often significantly bound to plasma proteins, and therefore, decrease in plasma protein concentration can increase the concentration of drug in the CNS. Perhaps the most important pharmacokinetic principle with respect to sedative infusions is the context-sensitive half-time, which is a calculated multicompartment model-derived graph of the time required for a 50% reduction in drug concentration at the effect site versus infusion duration for a constant-rate infusion. Some drugs (propofol) have a short half-time if the infusion is relatively short-lived (hours), but the half-time may increase significantly if the infusion is continued for days. The main difference in context-sensitive half-times between sedatives is the rate at which the half-times increase with infusion time. Compared with propofol and dexmedetomidine, benzodiazepines have significantly longer context-sensitive half-times at both short and long duration infusions—an important fact to consider when choosing a sedative.

Propofol

Propofol is a commonly used IV anesthetic induction agent that has excellent effectiveness in the ICU for varying levels of sedation. It works on a variety of CNS receptors to block neural transmission, including gamma-aminobutyric acid (GABA), and not only cause sedation, hypnosis, and amnesia, but also has anticonvulsant and antiemetic properties. It has no analgesic properties. It is highly lipid soluble and therefore crosses the blood-brain barrier and redistributes to vessel-poor compartments quickly, resulting in a fast-onset, fast-recovery when given in bolus doses and short infusions. It is, therefore, useful in spontaneous awakening trial (SAT) protocols (see “the case of lighter sedation” later) and in patients that require frequent neurological assessments. Given the context-sensitive half time, delayed awakening after prolonged infusions still occurs.

The most common adverse effect with propofol is hypotension caused by arterial and venous vasodilation in addition to direct myocardial depression. The hemodynamic effects are dose dependent, and are less profound with infusion rates targeted to moderate sedation than with bolus induction doses. The carrier solution (lipid solution containing egg lecithin and soybean oil) is potentially allergenic and has been implicated in the rare propofol infusion syndrome—a potentially devastating severe metabolic acidosis associated with hypotension, hypertriglyceridemia, rhabdomyolysis, ECG changes, and multiorgan dysfunction.

Dexmedetomidine

Dexmedetomidine is a potent α₂-adrenergic receptor agonist with a potency several fold higher than clonidine (α₂: α₁ activity = 1,620:1 and 220:1 for dexmedetomidine and clonidine, respectively). The relative densities of these receptors in the CNS account for the sedative, sympatholytic, and analgesic properties of the drug. Dexmedetomidine usually results in a lightly sedated, interactive, cooperative patient. Because of minimal respiratory depression, it is particularly useful in facilitation of liberation from mechanical ventilation and extubation, as an infusion can be safely continued during a spontaneous breathing trial through extubation. Although it is only Food and Drug Administration (FDA)-approved for ICU sedation at doses up to 0.7 μg/kg/h for 24 hours, there is empiric evidence to support safety at higher doses (1.5 μg/kg/h) for longer durations (up to 28 days).

Concerns with dexmedetomidine are primarily hemodynamic, as it can cause hypotension at rates similar to that of propofol. Loading doses of dexmedetomidine delivered too rapidly can cause severe life-threatening hypertension and bradycardia. Hypertension is likely caused by agonism of the peripheral α₁ or α_2B receptor subtype in the periphery resulting in vasoconstriction. Because of sympatholysis seen with continuous infusion, attempts to achieve deeper levels of sedation may be limited by bradycardia and hypotension. Lastly, it should be noted that the dosing of dexmedetomidine is unique among standard critical care infusions owing to its potency—μg/kg/h and not μg/kg/min.

Benzodiazepines

Benzodiazepines provide anxiolysis, sedation, hypnosis, and anticonvulsant effects through activation of GABA_A receptors in the brain. The most common benzodiazepines used in the ICU are midazolam and lorazepam, because they can be used effectively both as intermittent boluses and as infusions. However, the context-sensitive half-times of benzodiazepines are significantly longer than that of propofol or dexmedetomidine, and therefore, benzodiazepines may cause residual sedation hours to days after infusions are stopped. Because of hepatic metabolism, benzodiazepine clearance may be reduced in patients with hepatic dysfunction or when coadministered with cytochrome P₄₅₀ inhibitors.

Although midazolam has a shorter elimination half-life than other benzodiazepines, there is greater variability and potentially longer time to awakening with midazolam, likely secondary to accumulation of an active metabolite (conjugated and unconjugated α-hydroxymidazolam). This accumulation is especially important with concomitant renal dysfunction. Caution should be used in the elderly, both because benzodiazepines can cause paradoxical agitation and because altered pharmacokinetic factors such as increased volume of distribution and decreased elimination half-life often increase time to awakening.

Historical Context of Sedation for Mechanical Ventilation

Because the widespread implementation of positive pressure ventilation, use of sedatives to ensure patient-ventilator synchrony has been a mainstay of the practice of critical care. In the early days of mechanical ventilation, synchronized patient-ventilator interaction was very difficult or impossible, and was thought to require deep sedation to ensure optimal respiratory physiology. The patient experience of ICU care was considered to be inhumane and the provision of deep sedation was felt to provide comfort and amnesia to the experience. Treatment of ARDS involved normalizing respiratory physiology and gas exchange, which often involved deep sedation and comatose patients. In general, protocols to systematically assess the depth of sedation were not used. High infusion rates of intermediate-acting benzodiazepines and opioids were common, and patients were often unresponsive for days at a time. Although it is now apparent that this sedation paradigm directly contributes to worse outcomes, change in clinical practice has been slow. Even after a number of studies suggested that deep sedation may be harmful, 40% to 50% of patients were observed to be deeply sedated.¹⁵ Finally, strategies of ICU sedation have changed considerably; the evidence for these changes will be discussed here. As follows, the reader will find summaries of seminal works that have greatly informed current sedation paradigms, but are strongly encouraged to perform a more in-depth reading of the primary source material.

The Case for Lighter Sedation

Observations from more than 20 years ago noted that mechanically ventilated ICU patients sedated by continuous infusion had significantly longer duration of mechanical ventilation, ICU, and hospital lengths of stay than those who were sedated by intermittent intravenous boluses.¹⁶ Likely as a result, rates of documented ventilator-associated pneumonia (VAP) were also higher. In response, Kress and colleagues performed a trial comparing a protocol of daily sedative interruption (DSI, now referred to as SAT) to sedation as directed by the physician.¹⁷ About 150 patients were enrolled in the study at a single center, and the duration of mechanical ventilation, hospital LOS and ICU LOS were the primary endpoints. Duration of mechanical ventilation and ICU LOS were both decreased in the DSI group by 2.4 days (4.9 days, 95% CI: 2.5-8.6 vs 7.3 days, 95% CI: 3.4-16.1, P = 0.004) and 3.5 days (6.4 days, 95% CI: 3.9-12 vs 9.9 days 95% CI: 4.7-17.9, P = 0.02), respectively. There was no difference in rates of self-extubation or mortality. The authors concluded, “daily interruption of sedation is a safe and practical approach to treating patients who are receiving mechanical ventilation.”

Recognizing the results of this trial and appreciating the intertwined nature of mechanical ventilation and sedation, Girard et al published a multicenter study combining protocols of SATs (after a SAT safety screen) with spontaneous breathing trials (SBTs) and compared the protocol with usual care.¹⁸ The so-called “Awakening and Breathing Controlled Trial” (ABC trial) enrolled 336 patients in 4 centers with a primary outcome of time breathing without assistance during the first 28 days after enrollment. Not only did the protocol patients have an increased mean/median number of ventilator-free days of 14.7/20 versus 11.6/8.1 (P = 0.02 for mean), but secondary outcomes of 1 year mortality were significantly decreased in the protocol patients (44% vs 58%, P = 0.01). It is important to note that the rate of self-extubation was higher in the intervention group, but the rate of reintubation was not different.

Both the Kress et al trial and ABC trial had a control group with usual care, which usually consisted of sedation that was not interrupted. Although the Kress et al trial targeted sedation depth in both groups to achieve a score of 3 or 4 on the Ramsay sedation scale (3—responsive to commands and 4—asleep with a brisk response to a light glabellar tap or loud sound), patients in the control arm were only “awake” (responsive to commands) on 9% of patient-days. In the ABC trial, both control and intervention group had sedation that was titrated to different institution-specific endpoints. However, the control group had an increased duration of coma during the study period (3 days, IQR 1-7 vs 2 days, IQR 0-4, P = 0.002). Given these data, both control group patients were comatose for a significantly greater amount of time than the intervention group patients. An important question remained—would a SAT/SBT be superior to a strategy that simply titrated sedation to a lighter, noncomatose, level?

A multicenter trial attempted to answer this question by randomizing 430 patients to one of two protocols: (1) protocoled sedation targeting a RASS of –3 to 0 (sedated but opening eyes to voice [–3] to alert and calm [0]) plus daily sedation interruption or 2) protocoled sedation (RASS –3 to 0) alone without interruption of sedating medications.¹⁹ The primary outcome measure was time to successful extubation. There was no difference in time to successful extubation in the groups. Interestingly, secondary outcome measures showed an increase in the dose of sedating medications and the self-perceived nurse workload in the interruption group. There were no differences in self-extubation or delirium.

Given that sedation interruptions and perhaps lighter sedation improves outcomes, some ICUs have adopted a practice of no sedation for mechanical ventilation. In one such mixed medical-surgical ICU in Denmark where the nurse: patient ratio is 1:1 and extra staff are available to “verbally comfort and reassure the patient,” a RCT was performed to compare the intervention group (no sedatives) to the control group (sedation with goal Ramsay score 3-4 and daily sedation interruption).²⁰ Analgesia with as needed IV morphine was administered in both groups. The primary outcome measure was the number of days without mechanical ventilation in a 28-day period, with secondary endpoints being brain imaging (MRI or CT), accidental extubation, and VAP. The no sedation group had 4.2 days (95% CI: 0.3-8.1) more without mechanical ventilation in the 28 days after randomization (corrected for baseline variables). The secondary outcomes were not different between the groups. As assessed by the gold standard criteria, the DSM IV, more patients in the no sedation group experienced agitated delirium than in the control group (20% vs 7%, P = 0.04); these patients were treated with higher average doses of haloperidol, although the dose was small in both groups. This trial addressed the frequent appeals for a RCT comparing strategies of sedation versus no sedation and showed that, in carefully selected patients within a particular staffing model, sedation may contribute to prolonged mechanical ventilation and may be unnecessary. A large multicenter trial in Denmark and Norway (NONSEDA Study Group) is attempting to confirm these findings by enrolling 700 patients using the same protocol.

From currently available data, it is now apparent that (1) depth of sedation should be routinely monitored and quantified using a validated assessment tool, (2) titrating medications to keep patients continuously lightly sedated unless a contraindication exists (severe ARDS, refractory intracranial hypertension, status asthmaticus, or epilepticus) appears noninferior to standard therapy with DSIs, (3) titration of sedation and SBTs should be closely coordinated in a multidisciplinary team that involves respiratory therapists, nurses, and the medical team.

As the paradigm of ICU sedation continues to shift away from deep sedation and immobility, the list of accepted indications for use of NMBDs in mechanically ventilated patients has diminished considerably. However, NMBDs may still play an important role in management of critically ill patients in well-defined situations. These may include use to facilitate tracheal intubation, minimize systemic oxygen consumption in the setting of severe and refractory hypoxemia, improve outcomes in moderate and severe ARDS, treat shivering in patients undergoing targeted temperature management therapy, or treat refractory intracranial hypertension. Other “relative” indications, such as in protection of important vascular anastomoses (free flaps), taut wound closures and tracheal anastomoses, are dubious in nature and should be questioned. It is also important to note that administration of a NMBD excludes the use of subjective assessments of pain and sedation depth. In this setting, use of objective measures of brain function, such as bispectral index monitoring (BIS) or formal electroencephalography (EEG) to monitor depth of sedation should be considered.

NMBD Pharmacology

NMBDs have unique pharmacology that necessitates knowledge of the biochemistry of nicotinic acetylcholine receptors at the neuromuscular junction (NMJ). NMBDs are classified as either depolarizing or nondepolarizing based on their action at the postsynaptic nicotinic acetylcholine receptor.

Succinylcholine is the only depolarizing NMBD available in the United States—it depolarizes nicotinic acetylcholine receptors and either desensitizes the receptor, inactivates sodium channels and prevents propagation of an action potential, or a combination of both. It acts similarly to acetylcholine at the receptor, but binds significantly longer, and therefore, can cause significant hyperkalemia by the resulting efflux of potassium from muscle cells, especially in the setting of extrajunctional acetylcholine receptors. These extrajunctional receptors are expressed more frequently on muscle cells that have been denervated by neurologic disease (stroke, Guillian–Barre, spinal cord injury, severe burns, and prolonged immobility), and therefore, fatal hyperkalemia can result in these cases. Succinylcholine has other important adverse effects including hypertension, tachycardia, bradycardia, ventricular arrhythmias, and less commonly, increased intracranial pressure and malignant hyperthermia.

Nondepolarizing NMBDs are competitive inhibitors of acetylcholine, and inhibit depolarization at the NMJ. Therefore, the presence of extrajunctional acetylcholine receptors is not a concern with this class of NMBDs. These NMBDs are subclassified by duration of action (short, intermediate, and long) and chemical structure (benzylisoquinoline or aminosteroid). All commonly used nondepolarizing NMBDs are intermediate acting. A summary of the pharmacology of NMBDs commonly used in the ICU is presented in Table 16–5.

Neuromuscular Blockade for Facilitation of Mechanical Ventilation

NMBDs have been used to avoid patient-ventilator dyssynchrony in the setting of severe and/or refractory gas exchange abnormalities, particularly hypoxemia, but a number of case series suggest that they are associated with prolonged neuromuscular weakness. Although some of these cases were ascribed to decreased drug clearance in the setting of renal and hepatic dysfunction, some were thought to result from pathologic myopathy.²¹ This myopathy has also been demonstrated in rats after prolonged exposure to vecuronium. Many of the cases with potential myopathy involved aminosteroid NMBDs (see Table 16–5). Although there have been case reports of prolonged weakness after infusion of benzylisoquinoline, it is thought to be less of a concern than with aminosteroid compounds. The most commonly used benzylisoquinoline NMBD, cisatracurium, also has the advantage of being degraded by Hoffman elimination and ester hydrolysis in the plasma, and is therefore, not dependent on renal or hepatic function for clearance.

Infusion with cisatracurium for 48 hours for treatment of moderately severe and severe ARDS was recently studied in a double-blind RCT in France in 340 patients.²² The primary outcome measure (adjusted hazard ratio for death at 90 days) was 0.68 (95% CI: 0.48-0.98, P = 0.04) for the cisatracurium group, with no difference in ICU-acquired paresis. In both groups, a large number of patients were receiving corticosteroids for septic shock, which may contribute to ICU myopathy. A meta-analysis including this trial as well as two smaller studies also supports the notion that short-trem use of cisatracurium reduces hospital mortality and barotrauma without an increased risk of ICU-acquired weakness. Thus, current evidence supports the short-term use of cisatracurium for severe ARDS, and should be in any intensivist's armamentarium for treatment of severe ARDS. As mentioned previously, depth of sedation monitoring with BIS or EEG should be considered if NMBDs are used. However, NMBD use to facilitate mechanical ventilation in patients without moderately severe to severe ARDS is strongly discouraged.