Residual Paralysis after Emergence from Anesthesia

Absence of residual paralysis means that neuromuscular transmission has recovered sufficiently, and so the unaided patient can breathe normally, clear secretions, cough, prevent aspiration of gastric contents, and maintain a patent upper airway. Because this return to complete recovery cannot be assessed easily before emergence from anesthesia and even during the early postanesthesia recovery phase, anesthesiologists have to rely on surrogate measurements. With the introduction, in the early 1970s, of the train-of-four (TOF) stimulation applied to the ulnar nerve, it became necessary to correlate adductor pollicis response to indices of respiratory function (fig. 1). In a study conducted by Ali et al.  1on six healthy awake volunteers, vital capacity, inspiratory force, and expiratory force were found to be normal when TOF ratio (TOFR; the ratio of the fourth to the first twitch height) was more than or equal to 0.70.

Based on that evidence, the 0.7-TOFR threshold was considered to indicate adequate neuromuscular recovery for nearly two decades. However, in the 1990s, several lines of evidence indicated that clinically relevant neuromuscular block still persists at TOFR = 0.7. In human volunteers, hypoxic ventilatory drive was shown to be decreased by vecuronium up to a TOFR more than or equal to 0.9.2In another study, the ability to swallow was also found to be impaired when the TOFR was less than 0.9.3Masseter muscle function, assessed by the ability to hold a tongue depressor between one's teeth against resistance, did not return to normal unless TOFR equaled 0.8–0.9.4Therefore, a revisited TOFR threshold more than or equal to 0.90, obtained by force measurement or mechanomyography, was proposed in the late 1990s. With the advent of techniques measuring acceleration, or acceleromyography, a TOFR more than or equal to 1.0 was recommended.5 

The degree of residual paralysis can be evaluated in different ways: (1) clinical tests requiring the patient's cooperation, which normally can be performed only after emergence; (2) visual or tactile evaluation of responses to TOF or double-burst stimulation (DBS) at the adductor pollicis (qualitative or subjective assessment); and (3) measurement of the TOFR with a device (quantitative or objective measurement).

For the conscious and cooperative patient, several clinical tests have been proposed (table 1). Sustained head lift has been studied extensively and was found to correspond to maximum inspiratory pressures ranging from 50 to 53 cm H²O in unanesthetized volunteers partially paralyzed with d -tubocurarine.6However, in volunteers given subparalyzing doses of mivacurium, sustained head lift for 5 s correlated with a measured TOFR ranging from 0.45 to 0.75, lower than the recommended threshold of 0.9.4In patients, the sensitivity of the head-lift test was approximately 10%, whereas specificity was excellent at 87%,7which indicates that residual paralysis is likely in patients unable to maintain a sustained head lift. More recently, the ability to hold a tongue depressor between one's teeth despite the efforts of someone else to pull it out has been proposed as a more sensitive test.4Volunteers given mivacurium were unable to hold the tongue depressor at a mean TOFR less than 0.86, close to the 0.9 threshold. However, the sensitivity of the tongue-depressor test in patients (13%) was not much higher than that of the head-lift test, but its specificity was higher (90%).7When the head-lift or tongue-depressor tests are “passed,” the persistence of a certain degree of residual paralysis cannot be excluded, suggesting that more reliable tests are required. In addition, these clinical tests cannot be performed in the anesthetized patient.

Tests involving a stimulator and tactile or visual subjective assessment of the clinical observer have been devised (table 1). Several studies documented that visual or tactile evaluation of the TOF responses correlated poorly with measured TOFR.8–10Even experienced observers are unable to detect TOF fade visually or manually when the actual TOFR exceeds 0.4, which means that residual paralysis may be undetected if TOFR is in the range of 0.4 to 0.9.7This zone of blind paralysis can be reduced somewhat with the DBS mode of stimulation. With DBS, fade can be detected visually or manually up to a measured TOFR of 0.6, still well below the desired 0.9 threshold.9The failure of these subjective methods to detect residual paralysis was confirmed more recently.7,11The specificity of those two tests was good (98–99%), but sensitivity remained poor (11 and 14% for TOF and DBS stimulation, respectively).

Therefore, when fade is detected by tactile or visual means, a certain degree of residual paralysis can be expected with a high degree of certainty. In one study, residual paralysis (TOFR < 0.9) was present in 92–96% of subjects with demonstrated tactile or visual fade in response to TOF or DBS stimulation (positive predictive value).7However, complete recovery was seen in only half the patients with no fade (negative predictive value, 53–62%).7Hence, it is not surprising that the absence of clinical fade after DBS or TOF stimulation does not mean complete recovery.

Tetanic stimulation has also been used to evaluate residual paralysis. Assessing tetanic fade after a 50-Hz stimulation for 5 s does not provide any more information than TOF, because most observers do not detect tetanic fade when the TOFR is more than 0.4.12Tetanic fade after 100-Hz stimulation can be detected at a TOFR of 0.8–0.9, making it a more sensitive test.13However, this stimulation is very painful and must not be used on the awake patient. Also, it produces a posttetanic facilitation period of 5–10 min, during which the response to any test (TOF, DBS, or tetanus) is enhanced, spuriously indicating more recovery than is actually the case. Thus, neither the clinical tests applied after emergence nor the qualitative instrumental tests are sufficiently accurate to detect the absence of residual neuromuscular block.

Adequate neuromuscular recovery, defined as an adductor pollicis TOFR more than or equal to 0.90, requires the quantitative evaluation of TOFR, using measurement methods such as acceleration (acceleromyography), electromyography, force (mechanomyography), or displacement (kinemyography; table 1). To be clinically acceptable, these methods must have excellent reproducibility and be simple to use. For many years, mechanomyography at the adductor pollicis was the only technique available in the operating room and the PACU. The TOFR threshold of 0.9 was established with this device. However, mechanomyography instruments are cumbersome and difficult to set up, and so the technique has never gained wide clinical acceptance. Electromyography, which is based on the measurement of electrical activity in muscle, is easier to use and less cumbersome, but it is fragile, expensive, and subject to electrical interference from cautery.

With the introduction of acceleromyography monitors in the mid-1990s, the TOFR can now be quantified objectively in routine daily practice. These monitors are inexpensive, versatile, and relatively easy to set up. However, the limits of agreement are relatively wide12between data measured with this device and those obtained with the gold standard, the mechanomyography. The discrepancy between mechanomyography and acceleromyography is particularly important when TOFR is in the 0.9–1.0 range, because TOFR measured by acceleromyography tends to overshoot, displaying values more than 1.0. For example, when the mechanomyography TOFR reached 0.9 after atracurium administration, the corresponding acceleromyography TOFR ranged between 0.86 and 1.0 (mean 0.95).5The negative predictive values of acceleromyography TOFRs of 0.9, 0.95, and 1.0 to detect residual paralysis were 37, 70, and 97%, respectively. Therefore, to detect residual paralysis reliably with acceleromyography, recovery of a TOFR of 0.9 is considered insufficient, and a threshold of 1.0 is now recommended to confirm complete recovery from neuromuscular block.

Objective tests (table 1) of neuromuscular recovery can be applied to the awake patient in the PACU, but the response is not as reliable as in anesthetized subjects, because TOFR measurements can be affected by spontaneous movements of the thumb. Thus, the values obtained with two successive measurements may vary substantially. In one study,14the evoked thumb response was measured by acceleromyography after TOF stimulation in 253 patients after their arrival in the PACU. Current intensity was set at 30 mA, instead of the 50–70 mA commonly used in anesthetized subjects, to limit discomfort. Two TOF stimulations were applied successively and recorded at a 30-s interval. The first TOFR measurement indicated adequate neuromuscular recovery in 175 patients (TOFR ≥ 0.9), but for 40 of them, the second TOFR was less than 0.9. In the 78 patients considered to be partially paralyzed after the first measurement (first TOFR < 0.9), 21 of them had a second TOFR more than or equal to 0.9. In other words, the two TOFRs were discordant in 61 patients (24%). Based on that study, it can be concluded that two isolated acceleromyography TOFR do not accurately represent the patient's neuromuscular status and that repeated measurements (> 2) are needed.

The residual paralysis rate after emergence has been extensively evaluated during the last 30 yr with global frequencies ranging from 5 to more than 85%. This wide variability can be explained by substantial methodologic differences among those studies.

Residual paralysis was first documented in the late 1970s, when the threshold for neuromuscular recovery was considered to be a TOFR more than 0.7. It is not surprising that later studies based on a higher TOFR threshold detected a greater frequency of residual paralysis. For example, in a study published in 2003, 526 patients received a single intubating dose (2× the ED⁹⁵) of atracurium, vecuronium, or rocuronium. At the end of the procedure, which lasted 1–4 h, 16% had a TOFR less than 0.7 but as many as 45% had a TOFR less than 0.9.7In another study involving the same NMBDs (148 patients), the rate of residual paralysis reached 41% based on a TOFR value of 0.7 and 52% when 0.8 was considered as the threshold for recovery.15In patients receiving pancuronium, the frequency of residual paralysis, defined as a TOFR less than 0.7, was less (40%) than if defined as a TOFR less than 0.9 (85%).16With rocuronium, the percentages were lower, but the difference between the threshold of 0.7 (6%) and 0.9 (29%) persisted.16 

The longer the duration of NMBD action, the higher the frequency of residual paralysis, regardless of the TOFR threshold chosen.17In a nonrandomized study, residual paralysis, defined as a TOFR less than 0.7 in the PACU, was more frequent in patients given pancuronium (36%; 17/47) than in those who had received atracurium (4%; 2/46) or vecuronium (8%; 5/57).18 

It seems that patients receiving NMBDs by infusion are more likely to have residual paralysis. In 150 patients given atracurium or vecuronium, 100 received the NMBD as repeated boluses and 50 others by continuous infusion.19Neostigmine reversal was administered to 97% of cases. Residual paralysis, defined as a TOFR less than 0.7, was found in 12% of the bolus group patients and in 24% of the infusion group patients on arrival in the PACU. Fifteen minutes later, the problem persisted in 2 and 12%, respectively. These observations suggest that continuous infusion of NMBDs can increase the risk of residual paralysis at emergence.

The usefulness of intraoperative neuromuscular monitoring to reduce the frequency of residual paralysis on arrival in the PACU remains a matter of debate. The results of a recent meta-analysis indicated that the use of an intraoperative neuromuscular function monitor was not associated with a decrease of the residual paralysis rate.17However, that study included a number of uncontrolled trials. When analyzing only studies with adequate methodology, based on a Jadad score more than or equal to 3 (i.e. , at least a randomized controlled trial and description of withdrawals), only five articles met these criteria and four of them demonstrated a benefit of preoperative neuromuscular transmission monitoring to decrease the residual paralysis rate, whereas only one showed the opposite.10,20–22The usefulness of perioperative acceleromyography monitoring was evaluated between on two groups of patients given pancuronium and neostigmine reversal with monitoring (n = 19) or without (n = 21).21Acceleromyographic TOFR were measured for both groups after extubation. The TOFR was less than 0.7 for 11/21 (52%) unmonitored patients, whereas only 1/19 (5%) monitored had a TOFR less than this threshold.

A benefit of intraoperative monitoring was also found with intermediate-acting NMBDs. In a prospective, randomized, and double-blind study, the degree of residual paralysis after rocuronium use was compared between unmonitored and acceleromyography-monitored patients (80 patients in each group).20Residual muscle paralysis, defined as a TOFR less than 0.8, was found in 16.7% of the unmonitored and 3% of the monitored group. Therefore, the problem of residual paralysis can apparently be minimized by neuromuscular monitoring but cannot be definitively excluded. Finally, it was recently demonstrated in a randomized study that the incidence of incomplete neuromuscular recovery was less with quantitative (acceleromyography) than with qualitative (visual assessment of TOF) monitoring.23 

Obviously, reversal of nondepolarizing neuromuscular block seems to be one of the critical steps in reducing or to eliminating residual paralysis. To date, no prospective, randomized, and double-blinded studies have compared the rates of residual paralysis between two therapeutic arms: reversal versus  placebo. Therefore, the efficacies of different reversal strategies can only be assessed indirectly by analysis of the existing literature.

In one study involving 148 patients receiving intermediate-duration nondepolarizing NMBDs, 101 received no reversal, whereas the remaining 74 patients received neostigmine, but the allocation was not randomized.15Residual paralysis defined as a TOFR of less than 0.8 was found in 48% of patients who received neostigmine compared with 59% in those who did not. The difference was not statistically different. Indirect evidence of the efficacy of reversal agents can be estimated by comparing the results of Bevan et al.  18with those obtained by Baillard et al.  24In the former study, 58% of 103 anesthetized patients paralyzed with either atracurium or vecuronium were given either neostigmine or edrophonium reversal, and the frequency of residual paralysis, defined as a TOFR less than 0.7, was 7%.18In the latter study, 568 patients received vecuronium, and no subsequent reversal, and residual paralysis, defined with the same criteria, was detected in 42% of subjects.24 

The physiologic consequences of residual paralysis, such as respiratory impairment,1,2upper airway collapse,4,6and abnormal swallowing,3are well documented, but the vast majority of the studies were conducted on nonanesthetized healthy volunteers in controlled conditions. However, a direct link between such residual paralysis and poorer outcome is difficult to establish, because many other factors can contribute to respiratory complications after a procedure (i.e. , residual effects of other anesthetic agents, type of procedure, comorbidities, and duration of the procedure). However, the adverse effects of residual paralysis after emergence have been documented in clinical studies (table 2). Some of them demonstrated poorer outcomes for anesthetized patients in terms of postoperative morbidity and mortality when residual paralysis persisted.

Thus, residual paralysis is an unwanted side effect in the immediate postoperative period and a risk factor for respiratory complications. That risk could be decreased by the systematic use of neuromuscular monitoring and judicious administration of reversal agents.

Lunn et al.  25demonstrated 25 yr ago in a survey based on anonymous reporting of deaths within the first 6 days after anesthesia that 11/32 (34%) deaths attributed entirely to anesthesia were caused, at least in part, by postoperative respiratory failure. Residual paralysis was considered to be contributory in six of those deaths (table 2). During the same decade, Cooper et al.  26reported on the causes of unexpected admission to the intensive care unit because of a complication of anesthesia during a 5-yr period. There were 53 cases, and the mortality rate was 17%. The majority (33 of 53) of complications occurred in the recovery period. Twenty-four of these 33 cases were due to ventilatory inadequacy after reversal of neuromuscular block. Tiret et al.  27conducted a French national survey on anesthesia-associated mortality by retrospectively analyzing 200,000 anesthesia procedures and found that half of the 67 anesthesia-associated deaths resulted from postanesthesia respiratory depression. More recently, a list of risk factors, directly related to anesthesia management and considered to be responsible for postoperative mortality and severe morbidity detected in the first 24 h, were identified.28Among them, omitting to reverse a residual block was associated with a 10-fold increased risk for death or coma. This finding provides indirect evidence that residual paralysis could be implicated in death and severe morbidity.

Direct evidence of morbidity associated with residual paralysis during emergence has been demonstrated after pancuronium administration. In a randomized study, the frequency of residual paralysis, defined as a TOFR of less than 0.9, was significantly higher in patients given pancuronium (85%) than those administered rocuronium (29%).16Hypoxemia (defined as Spo²< 93%) in the PACU was found more frequently in patients who had received pancuronium. An association between residual paralysis (TOFR < 0.9) and postoperative hypoxemia was demonstrated.29According to a large controlled study on 693 patients randomized to receive pancuronium, vecuronium, or atracurium for abdominal, gynecological, or orthopedic surgery, respectively, a potential risk factor for development of postoperative pulmonary complications, defined as atelectasis on chest x-ray 2 days after surgery, was identified as a TOFR less than 0.7 on arrival in the PACU after pancuronium administration.30Significantly, more of those patients with residual paralysis developed postoperative pulmonary complications (17%; 10/59) when compared with patients without such residual paralysis (5%; 8/167). These findings demonstrate that residual paralysis on arrival in the PACU increases the risk of subsequent pulmonary morbidity.

To summarize, the evidence suggests that undetected residual paralysis during emergence from anesthesia is common and may have deleterious clinical consequences. Although detection and treatment of residual paralysis are achieved with neuromuscular monitoring and/or reversal of block, surveys have shown that adherence to these principles is relatively poor. For example, a national survey conducted in France showed that 50% of anesthesiologists never use a peripheral nerve stimulator, and only 32% systematically or frequently administered an acetylcholinesterase inhibitor31when an NMBD had been given. Thus, residual paralysis occurs probably more frequently in actual practice than in studies where monitoring and the use of reversal agents were standardized. It follows that it is essential to change clinicians' approach to the management of residual paralysis, and this constitutes a safety issue. To avoid residual paralysis, we must focus on management during anesthesia. No one-size-fits-all solutions are available, but a number of strategies can be applied depending on the type of procedure and the patient's medical status.

A nondepolarizing NMBD should not be given when the procedure can be performed without paralysis and the airway secured using a supraglottic device, such as a laryngeal mask airway. However, NMBDs improve the quality and ease of tracheal intubation and lead to less subsequent laryngeal morbidity. Thus, neuromuscular block is recommended for tracheal intubation, even if relaxation is not required for surgery. If the duration of the procedure is short, succinylcholine can be an alternative to nondepolarizing NMBDs but exposes the patient to that drug's adverse effects.

Some authors have studied the usefulness of a low dose nondepolarizing NMBD (< 2 × the ED⁹⁵) to improve the quality of tracheal intubation,32but the impact of this strategy on the frequency of residual paralysis at the end of the procedure has never been evaluated. When a high dose (i.e. , 2 × the ED⁹⁵) of intermediate-acting nondepolarizing NMBD, for example, rocuronium, atracurium, and vecuronium, is administered to facilitate tracheal intubation, clinicians must be aware that an interval exceeding 2 h from NMBD injection to the arrival in the PACU does not guarantee an absence of residual paralysis,7highlighting the need, before extubation, for neuromuscular monitoring and reversal, if needed.

When muscle paralysis is necessary during a procedure, the choice of the drug is based on the planned duration and the patient's medical status. Regardless of the duration of the procedure, long-acting NMBDs, such as pancuronium, should be avoided because the residual paralysis rate on arrival in the PACU is particularly high.16–18,30For short procedures, mivacurium can be a valuable option. However, mivacurium is no longer available in North America.

Because the steroidal compounds are eliminated via  liver and/or kidney, benzylisoquinolines, such as atracurium or cis -atracurium, seem to be a better choice for patients with liver or renal insufficiency. Bolus administration of NMDBs should be preferred to infusions because the residual paralysis rate is higher with the latter.19Halogenated agents lower the dose of the nondepolarizing NMBD required and prolong their duration of action. To decrease the frequency of residual paralysis, intravenous anesthesia could theoretically be more appropriate than inhaled anesthesia. However, the frequency of residual paralysis associated with these two anesthesia regimens has never been compared in a well-designed study.

Prevention of hypothermia is essential because it increases the duration action of NMBDs. Although the usefulness of intraoperative neuromuscular monitoring to lower the frequency of residual remains a mater of debate,17it seems easier and more convenient to use a nerve stimulator to adjust the degree of block during the procedure, to detect residual paralysis during emergence and assess the need for reversal.

During emergence, the focus should be on preventing residual paralysis. Two options are available: (1) allowing spontaneous recovery or (2) reversing neuromuscular block with an acetylcholinesterase inhibitor or selective binding relaxant.

If spontaneous recovery is chosen, there should be solid evidence that neuromuscular function has returned to a TOFR of more than or equal to 0.9 before tracheal extubation. As discussed earlier, none of the traditional clinical tests and qualitative neuromuscular tests can accurately and reliably indicate a return to a TOFR of more than or equal to 0.9. Hence, it is easier and more convenient to use objective monitoring, such as an acceleromyography device, if reversal is to be omitted. Time is not a guarantee of recovery: residual paralysis can persist more than or equal to 4 h after an intubating dose of rocuronium, vecuronium, or atracurium is given.7 

After administration of a nondepolarizing NMBD, a neuromuscular monitoring device helps greatly in deciding whether to inject a reversal agent: acetylcholinesterase inhibitors or sugammadex. Acetylcholinesterase inhibitors block the enzyme acetylcholinesterase, which normally hydrolyzes acetylcholine at the neuromuscular junction. As a result, more acetylcholine competes with the nondepolarizing NMBD for access to the receptors, so that some neuromuscular function is restored. However, the efficacy of acetylcholinesterase inhibitors is limited, because their maximum effect is reached when enzyme inhibition approaches 100%. Clinically, this ceiling effect is probably reached at neostigmine doses of 0.04–0.07 mg/kg or equivalent. This phenomenon implies that acetylcholinesterase inhibitors are not effective when the block is too intense. Therefore, it is essential to wait until some degree of spontaneous recovery has been achieved before administering the acetylcholinesterase inhibitor. Time to adequate recovery (TOF ≥ 0.9) declines from a median of 22 min, when if only one twitch is visible, to 16 min when four twitches are visible.33Thus, it is now recommended to wait until four twitches are visible before giving neostigmine (fig. 2A). Giving higher doses when block is intense is not effective because of the ceiling effect.

Acetylcholinesterase inhibitors also have cholinergic effects, notably bradycardia, and increased volumes of salivary and bronchial secretions. These actions can be counteracted by anticholinergic drugs, such as atropine or glycopyrrolate. However, administration of the mixture is associated with an increased frequency of arrhythmias.

Another approach can be proposed to obtain block reversal. Instead of providing more acetylcholine at the neuromuscular junction, a substance that binds selectively to the NMBD has been developed. This drug, called sugammadex, is a γ-cyclodextrin, which is a ring composed of sugars, that selectively binds rocuronium and has a somewhat weaker affinity for vecuronium and pancuronium. Sugammadex is an abbreviation of sugar and γ-cyclodextrine. It does not bind to other classes of NMBDs, such as succinylcholine, atracurium, cis -atracurium, and doxacurium. At the time of writing, sugammadex had been approved for marketing in Europe, Australia but not in North America.

The main advantages of sugammadex are its rapid speed of recovery with minimal interindividual variation regardless of block level and finally lack of cholinergic side effects. In addition, it achieves recovery to TOFR more than or equal to 0.9 rapidly (3–5 min), when the appropriate dose is given. However, the sugammadex dose required depends on intensity of the block. When two twitches are visible after a TOF stimulation at the ulnar nerve, 2 mg/kg should suffice.34For more profound blocks, at reappearance of a posttetanic count of 1–2 at the adductor pollicis, 4 mg/kg is usually needed,35and as much as 8–16 mg/kg can be required if sugammadex is given a few minutes after an intubating dose of rocuronium (0.9–1.2 mg/kg; fig. 2B).36With vecuronium block, the sugammadex dose is approximately the same as with rocuronium.37No data are available for pancuronium.

Sugammadex is not effective against other NMBDs. Because the sugammadex dose depends on the level of block, it is strongly recommended to monitor neuromuscular function before and after its administration to determine the dose and evaluate its efficacy. Regardless of the strategy selected (spontaneous recovery or reversal), measured TOFR more than or equal to 0.9 is required before tracheal extubation.

Residual paralysis is an anesthetic complication that can be avoided by careful management. As an example, a 10-yr survey in a single hospital demonstrated that a well-implemented strategy based on promotion of neuromuscular monitoring and reversal led to a dramatically decreased residual paralysis rate in the PACU.38From 1995 to 2004, patients receiving intermediate-acting NMBDs were prospectively studied during 3-month period in 1995 (n = 435), 2000 (n = 130), 2002 (n = 101), and 2004 (n = 218). In 1995, quantitative measurement of neuromuscular block was performed in only 2% of cases, and 6% of patients received reversal agents. In 2005, corresponding figures were 60 and 42%, respectively. During the same time period, the frequency of residual paralysis, defined as a TOFR less than 0.9, decreased from 63 to only 3%, demonstrating that the systematic application of simple measures can make a difference.