Guidelines on muscle relaxants and reversal in anaesthesia adductor pollicis *, **

Please cite this article as: Plaud B, Baillard C, Bourgain J-Louis, Bouroche G, Desplanque L, Devys J-Michel, Fletcher D, Fuchs-Buder T, Lebuffe G, Meistelman C, Motamed C, Raft J, Servin F, Sirieix D, Slim K, Velly L, Verdonk F, Debaene B, Guidelines on muscle relaxants and reversal in anaesthesia adductor pollicis *, **, Anaesthesia Critical Care and Pain Medicine (2020), doi: https://doi.org/10.1016/j.accpm.2020.01.005

J o u r n a l P r e -p r o o f 13 If the experts did not have enough data from the literature to allow them making a recommendation, an expert opinion was then proposed and, if at least 70% of the experts agreed with the proposal, it was approved.
The summaries prepared by the experts and application of the GRADE® system resulted in twenty-nine recommendations being made. Among the formalised recommendations, eleven have a high level of evidence (GRADE 1±) and nineteen have a low level of evidence (GRADE 2±). For five recommendations, the GRADE® method could not be applied, and expert advice was provided for two of them. After two rounds of scoring and an amendment, strong agreement was reached for all the recommendations. Safety is provided by the use of oxygenation methods immediately available prior to the onset of desaturation (SpO2 < 95%). Regardless of the induction protocol, with or without a muscle relaxant, the safety margin between apnoea duration and O2 reserves is small. This situation is all the more common with a high body mass index and high O2 consumption. Testing mask ventilation quality before injecting a muscle relaxant increases the duration of induction by adding the time it takes for the muscle relaxant to work to that of the other anaesthetic agents.
Testing the quality of mask ventilation before administering a muscle relaxant may be proposed bearing in mind that: 1) injecting a muscle relaxant may lengthen the duration of apnoea and induce desaturation due to the inability to ventilate with the mask, and 2) it must be possible to wake up the patient at any time if the situation so warrants.
The need to wake up a patient due to the inability to intubate is extremely rare and poorly documented: there have been only two cases out of 100 difficult intubations among 11,257 intubations [14] and nine cases among 698 patients with both difficult intubation and mask ventilation [15]. Rapid recovery of spontaneous ventilation after induction does not prevent the occurrence of oxygen desaturation [16], which justifies maintaining a high O2 reserve at all times. Neuromuscular blockade does not necessarily increase apnoea duration and risk of desaturation. Following administration of thiopental and suxamethonium in healthy volunteers, four out of twelve subjects showed a decrease in oxygen saturation to below 80% before recovery of neuromuscular blockade. Among twenty-four healthy volunteers, injection of propofol 2.0 mg/kg and remifentanil 1.5 or 2.0 µg/kg resulted in profound desaturation in five subjects (four in the 2.0 µg/kg group and one in the 1.5 µg/kg group) [17]. Increasing the dose of remifentanil from 1.0 to 2.0 µg/kg increased apnoea time substantially (270 s to 487 Page 16 of 75 J o u r n a l P r e -p r o o f 16 s) while around 10% of patients had intubation conditions deemed unacceptable [18]. Using remifentanil at a dose of 4 µg/kg provides intubation conditions comparable to those observed after suxamethonium at the expense of blood pressure lowering, and an increase in duration of apnoea (12.8 min vs. 6.0 min) [19]. All publications report the use of propofol at doses between 2.0 and 2.5 mg/kg. The return to consciousness judged on the response to simple orders and the bi-spectral index (BIS) is 529 s (standard deviation 176 s) after injecting 2 mg/kg of propofol [20]. Nevertheless, awakening does not imply disappearance of the effects of propofol on the laryngeal muscles; the latter remain measurable for concentrations of as low as 0.7 µg/ml [21]. The effect of suxamethonium has been evaluated as lasting seven to eight minutes long using laryngeal and diaphragmatic electromyography [22] and about twelve minutes using accelerometery [23], with significant variability in both cases.
Reducing the dose of suxamethonium does not significantly shorten this apnoea time [24].
Use of sugammadex after rocuronium results in a faster and more reliable recovery time than use of suxamethonium [23], if the right number of vials of sugammadex is immediately available [25]. Injecting sugammadex after rocuronium gives better results than suxamethonium in terms of mean value (4.7 min) and individual variability [23].
Neuromuscular blockade reversal using sugammadex may be ineffective and lead to a situation requiring emergency tracheal access [26].
Does a muscle relaxant impair mask ventilation in patients without difficult mask ventilation criteria and/or difficult intubation? Muscle relaxant administration improves muscle relaxation especially when using high doses of sufentanil [27] or remifentanil [28] or low doses of a hypnotic [29]. Five studies investigated the effect of muscle relaxant injection on respiratory mechanics, and all used different criteria. Two of these included a small number of patients (30 and 32 respectively) and reported non-significant [30] or slightly significant results in favour of suxamethonium [31]. Interindividual variability in muscle relaxant response is important, suggesting that some patients are already fully relaxed before paralysis. In two studies (125 patients, some of whom were obese [32], and 210 patients [33]), the improvement in ventilation after neuromuscular blockade was significant, although impairment was seen in 19% of them [33]. Improved mask ventilation was also brought to light on objective criteria after injecting rocuronium [34].
Based on a database of four hospitals including 492,239 general anaesthesias, the combination of mask ventilation difficulty and intubation difficulty was reported in 698 patients (0.4%) [15]. Improved mask ventilation after neuromuscular blockade was observed for 19 patients, with no worsening. In a prospective series of 12,225 patients with difficult intubation criteria, improved ventilation after muscle relaxant injection was noticed in 56 out of 90 patients presenting difficult mask ventilation (with no impairment for the others) [35].
Question 2: Is the use of muscle relaxants necessary to facilitate tracheal intubation? PICO. P = "adult patients receiving muscle relaxants for a scheduled procedure with tracheal intubation", I = "muscle relaxants (muscle relaxant type: suxamethonium, atracurium, cisatracurium, mivacurium, pancuronium, rapacuronium, vecuronium and rocuronium", C = "no muscle relaxants", O = either positive, for "assessment of intubation conditions, either by the criteria of Cormack and Lehane [36,37] or those of the "Good Clinical Research Practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents" [38], or negative, for "upper respiratory discomfort: sore throat, hoarseness, vocal cord injury, pharyngeal injury, dental injury, severe complications: tracheal perforation, oesophageal intubation, inhalation lung disease, allergic reaction." R2.1 -The use of a muscle relaxant is recommended to facilitate tracheal intubation.

Rationale
Since the 1999 consensus conference, several studies have examined the issue of whether muscle relaxants facilitate tracheal intubation. To this end, thirty-two randomised and controlled studies, a cohort study and a systematic review were analysed . The muscle relaxants currently available in France (from the oldest to the most recent) are suxamethonium, atracurium, rocuronium, mivacurium and cisatracurium. Studies involving doses of muscle relaxant below the 95% effective dose (ED95) at the adductor pollicis were not considered. If several doses of the same muscle relaxant were used per study, the dose closest to twice the ED95 was retained. For those studies comprising several protocols without a muscle relaxant, a second analysis was conducted considering only the best protocol. This is how data from 1,247 patients who had received a muscle relaxant and 1,422 who had not received a muscle relaxant (best protocol analysis: 1,092 patients) were analysed for a total of 2,669 patients (best protocol analysis: 2,339 patients). Without a muscle relaxant, 350 patients (best protocol analysis: 283 patients) presented poor intubating conditions. This corresponds to a rate of 24.6% (best protocol analysis: 25.9%). With a muscle relaxant, 51 patients (4.1%) had poor intubating conditions. The cohort involved 103,784 patients, of whom 28,201 were intubated without and 75,583 with a muscle relaxant. Without a muscle relaxant, the rate of poor intubating conditions was 6.7% vs. 4.5% using a muscle relaxant.
This study identified muscle relaxant-free intubation as an independent risk factor for difficult intubation. As a result, these data support the concept that, compared with a muscle relaxant-free protocol, the use of a muscle relaxant facilitates tracheal intubation.
R2.2 -The use of a muscle relaxant is recommended to reduce pharyngeal and/or laryngeal injury.

Rationale
The quality of intubation can have clinical consequences for the patient. Mencke et al. [57] established a link between intubation quality and postoperative complications such as vocal cord damage and hoarseness. Six randomised controlled studies in 746 patients (best protocol analysis: 694 patients) were analysed [42,43,45,57,67,72]. According to these studies, 90 patients (best protocol analysis: 65 patients) sustained pharyngeal and/or laryngeal injuries. This corresponds to a rate of 22.6% (best protocol analysis: 18.7%). By using a protocol with a muscle relaxant, this incidence could be reduced to 9.7%. This reduction in pharyngeal and/or laryngeal lesions was seen in all six studies analysed. The absence of a muscle relaxant to facilitate intubation increases the risk of upper respiratory injury or discomfort. While a pharyngeal and/or laryngeal injury rate of nearly 10% despite good intubating conditions may be surprising, these complications may be caused by several factors other than intubating conditions. The most important are certainly the tracheal tube size and extubation conditions. Suxamethonium provides excellent intubation conditions more frequently than rocuronium (RR = 0.86; 95% confidence interval 0.81-0.92; I² = 72%). In a subgroup analysis comparing suxamethonium 1.0 mg/kg with rocuronium at a dose greater than 0.9 mg/kg, no superiority of suxamethonium was found. However, the heterogeneity noted between the different studies warrants a GRADE 2 level for the recommendation. Without muscle relaxant, the success rate of laryngeal mask insertion is commonly high and ventilation conditions are often satisfactory [73,74]. Neuromuscular blockade is probably useful when inserting a supraglottic device when the doses of hypnotic and opioid agents used for induction are low [75,76]. Anaesthesia protocols that exclude propofol as an induction agent have a high rate of adverse insertion events whose incidence decreases with neuromuscular blockade [77,78]

Rationale
In the event of upper airway obstruction, administering a muscle relaxant is proposed in the same way as a change in supraglottic device size or adjustment of position [79]. Two clinical entities must be distinguished: glottic closure resulting in incomplete or easily reversible obstruction, and laryngospasm involving complete glottic closure which is irreducible using the standard methods of ventilation [80]. Muscle relaxation is highly recommended during laryngospasm even if injection of propofol (0.25 to 0.8 mg/kg) is effective in the majority of cases (77%) [81]. In the absence of a muscle relaxant, muscle relaxation under general anaesthesia is not always complete. Moreover, opioids tend to increase muscle tone [82]. In the event of glottic closure not related to laryngospasm, the administration of a muscle relaxant is useful after providing sufficiently deep anaesthesia. The best agent is suxamethonium, which is effective in all cases [83]. In addition, it is fast acting, and the  [84]. It is most often given intravenously (1.0 mg/kg) but intramuscular and even sublingual administrations have been proposed (4.0 mg/kg) [85,86]. In children under three years of age, atropine (0.02 mg/kg) it is usually associated to avoid bradycardia or even cardiac arrest [83]. It is essential to be aware of the adverse effects of suxamethonium and to comply with the contraindications of the drug [85,87]. Comparable efficacy may possibly be achieved with a non-depolarising muscle relaxant because low concentrations of muscle relaxant are required to relax the laryngeal muscles [87]. A low dose is sufficient to achieve glottic opening if the depth of anaesthesia is adequate (rocuronium or atracurium 0.1 to 0.2 mg/kg) [82,88]. In all cases, reduction of complications associated with upper airway obstruction is optimised when it is integrated within a quality approach. The immediate availability of suxamethonium and atropine, especially in paediatric operating theatres, is crucial as when combined with neuromuscular blockade for intubation, it reduces the incidence of cardiac arrest and severe airway obstruction accidents nearly by half [89].

Question 4: Is it necessary to monitor neuromuscular blockade for airway management?
PICO. P = "adult patients receiving non-depolarising muscle relaxants for a scheduled procedure with tracheal intubation", I = "muscle relaxants (muscle relaxant type: atracurium and rocuronium)", C = "no monitoring", O = "assessment of intubation conditions, using either the Cormack & Lehane criteria [36,37] or those of Good Clinical Research Practice (GCRP) in Pharmacodynamic Studies of Neuromuscular Blocking Agents" [38]. There are no studies supporting a recommendation concerning the relevance of neuromuscular monitoring for intubation compared to a fixed time following injection of rocuronium or atracurium.
R4.1 -The experts suggest that if instrumental neuromuscular blockade monitoring is used, the selected site should be the corrugator supercilii muscle because of its sensitivity to muscle relaxants and its kinetics of neuromuscular blockade, which are comparable to those of laryngeal muscle. Experts' opinion.

Rationale
Intubating conditions are worse when the orbicularis oculi is used to decide when to perform tracheal intubation compared to when the corrugator supercilii or the adductor pollicis are used [90,91]. Poor intubation conditions are only found in the group of patients with orbicularis oculi monitoring. The onset time of maximum neuromuscular blockade at the corrugator supercilii is comparable to that of the laryngeal adductor muscles, whereas that of the orbicularis oculi corresponds much more closely to those of muscles which are more sensitive to muscle relaxants (e.g. adductor pollicis) [92]. Under the same intubating conditions, corrugator supercilii monitoring reduces the time taken to achieve tracheal intubation compared to adductor pollicis monitoring [93,94]. These latter two studies indicate the use of the orbicularis oculi instead of the corrugator supercilii. In fact it was the corrugator supercilii that was assessed. With atracurium or rocuronium administered at a suitable dose for tracheal intubation, the fact of systematically waiting for the mean onset time to elapse before starting laryngoscopy ensures the best conditions for intubation. shown no benefits for the deep neuromuscular blockade compared to the moderate blockade [106,107]. However, no trials have shown any difference between the deep blockade and the moderate blockade in terms of intraoperative adverse surgical events or specific morbidity associated with poor exposure. Nevertheless, this lack of difference should be interpreted with caution as the trials all had small sample sizes (n = 24 to 102), which were insufficient to reveal any significant difference. It is therefore impossible to make a recommendation on the depth of neuromuscular blockade required to achieve a reduction in intraoperative and postoperative surgical morbidity. The experts only found a few studies focusing on the reduction of pneumoperitoneum pressure according to the depth of neuromuscular blockade: moderate (1 to 2 TOF responses at the adductor pollicis) or deep (1-2 post-tetanic count (PTC) responses at the adductor pollicis). The randomised study by Madsen et al. (n = 14) showed that a deep blockade increased the distance between the umbilical trocar and the promontory by 0.3 cm. This difference is not clinically significant, and the study did not evaluate the risk associated with inserting the first trocar [100]. Three randomised prospective studies were retained. A first randomised study of 67 patients showed a decrease in insufflation pressure for the deepest neuromuscular blockades. The results remain difficult to interpret as the authors compared two levels of deep blockade by monitoring the corrugator supercilii but without a group under moderate blockade, i.e. 1 to 2 TOF responses at the adductor pollicis [103]. In another prospective study including 61 patients, the intraabdominal pressures required to ensure satisfactory surgical conditions during colorectal surgery were significantly lower in the deep neuromuscular blockade group (9 mmHg) than in the moderate blockade group (12 mmHg, p < 0.001) [104]. The third prospective study involved 62 patients undergoing cholecystectomy with an initial insufflation pressure of 8 mmHg. The results showed that insufflation pressure needed to be increased to 12 mmHg in 34% of cases in the moderate neuromuscular blockade group versus 12% in the deep blockade group (p < 0.05) [105]. A recent open prospective study showed that the deep neuromuscular blockade allowed a 25% reduction in intra-abdominal pressure compared to an absence of neuromuscular blockade [108].

Rationale
The SFAR recommends intraoperative monitoring of neuromuscular blockade throughout anaesthesia when a muscle relaxant is administered. This recommendation applies both in the operating theatre and the post-anaesthesia care unit. Monitoring the neuromuscular blockade with a nerve stimulator is more accurate than clinical assessment [109]. It is recommended to stimulate the ulnar nerve on the wrist with a visual or tactile evaluation of contraction of the adductor pollicis, given the ease of access and the possibility of quantifying the response in this muscle [110]. The gold standard for intraoperative stimulation remains the evaluation of the adductor pollicis response to train-of-four (TOF) stimulation at the ulnar nerve. The presence of one to two responses after TOF stimulation at the adductor pollicis indicates recovery of about 10% of initial muscle strength [109]. When a deep neuromuscular blockade of the body's most resistant muscles is indicated (diaphragm, abdominal wall muscles), it is recommended to wait until the four responses to TOF stimulation at the adductor pollicis have disappeared and to monitor via post-tetanic count (PTC) stimulation [101,111]. In this case, the presence of one or two responses at the adductor pollicis indicates complete paralysis of the abdominal muscles [111]. TOF stimulation of the facial nerve and visual evaluation of response in the corrugator supercilii offers an alternative to PTC. The neuromuscular blockade profile of the corrugator supercilii is comparable to that of more resistant muscles such as the laryngeal adductor muscles and the diaphragm. Monitoring the orbicularis oculi after stimulation of the facial nerve provides information comparable to that obtained with the adductor pollicis [90].
This may be useful in the absence of intraoperative access to the upper limbs. At the end of the Page 28 of 75 J o u r n a l P r e -p r o o f 28 procedure, it is recommended to switch over to adductor pollicis monitoring as soon as possible to quantify recovery [112]. Instrumental acceleromyographic monitoring of the adductor pollicis is more accurate than simple visual or tactile evaluation of muscle contractions [110].

Rationale
No clinical test is sensitive enough to detect a residual neuromuscular blockade [113].
Qualitative measurement of the ratio between the fourth and first TOF response (T4/T1) of ≥ 0.9 is required to eliminate a diagnosis of residual neuromuscular blockade [114]. Not all muscles are equally sensitive to the effects of muscle relaxants. Monitoring should be carried out on a muscle with high sensitivity to muscle relaxants and slow recovery kinetics. The adductor pollicis meets this profile and is recommended [112]. Only quantitative instrumental J o u r n a l P r e -p r o o f 29 monitoring using the T4/T1 ratio measurement at the adductor pollicis with supramaximal stimulation of the ulnar nerve can be used to assess residual neuromuscular blockade [115,116]. The consequences of a residual neuromuscular blockade and the absence of reversal are higher morbidity and mortality within the first twenty-four hours postoperatively [117], a greater risk of critical respiratory events in the recovery room [118,119], a greater risk of postoperative pneumonia [120,121], a greater risk of pharyngeal muscle dysfunction [122], and delayed discharge from the recovery room [123]. The degree of optimal neuromuscular blockade prior to neostigmine administration was determined in two studies. In the first prospective, randomised study [124], sixty-three patients were divided into four groups based on the number of tactile responses of the Therefore, the degree of blockade prior to reversal by neostigmine must have spontaneously recovered at least four visual or tactile TOF responses. The second study, of similar methodology, analysed the time it took to obtain a TOF ratio of > 0.9 after rocuronium had been administered [125]. A hundred and sixty patients were divided into eight groups of twenty according to the number of TOF responses at the adductor pollicis (one, two, three and four responses) and according to the anaesthesia maintenance agent (propofol or sevoflurane). Neostigmine was administered at a dose of 70 µg/kg. The time taken to achieve a TOF ratio > 0.9 in the propofol groups (minutes, median and range) was 8.6 (4.7-18.9), 7.5 pollicis responses to TOF stimulation caused a complete reversal in less than ten minutes. On the other hand, under sevoflurane anaesthesia, reversal was not complete within ten minutes in all patients. All these results suggest that four responses to TOF, corresponding to a measured TOF ratio of ≥ 0.2, is the minimum to obtain before administering neostigmine.
R7.3 -It is recommended to administer neostigmine with neuromuscular blockade monitoring at the adductor pollicis, at a dose between 40 and 50 µg/kg adapted to ideal body weight, but not to increase the dose beyond this level, and not to administer it in the absence of residual blockade (Fig. 1).

Rationale
The dose-response relationship of neostigmine administered during a deep blockade, defined by a T1/T0 ratio of 0.01, and during a more moderate blockade, defined by a T1/T0 ratio of 0.1, was established after a neuromuscular blockade by atracurium (T1 is the value of the contraction of the first response to TOF and T0 the value of the first response to TOF before the neuromuscular blockade) [126]. The dose of neostigmine required to obtain a TOF ratio > 0.7, 10 minutes after reversal at T1/T0 = 0.1 (moderate blockade) was 50 ± 7 µg/kg (mean ± SD) and 49 ± 6 µg/kg at T1/T0 = 0.01 (deep blockade) (not significant). It thus appears that, under the study conditions, the dose of neostigmine was in the range of 40 to 50 µg/kg. These results were confirmed when rocuronium was administered [127].

Is there any point in increasing the dose of neostigmine?
Three doses of neostigmine (20, 40 and 80 µg/kg) were administered after randomisation in twenty-seven adult patients when the atracurium-induced blockade had spontaneously returned to a T1/T0 ratio value of between 0.05 and 0.1 [128]. The time it took to obtain a
In a recent study, three doses of neostigmine (10, 20 and 40 µg/kg) were compared to a placebo when the blockade induced by cisatracurium or rocuronium was spontaneously reversed, reaching 0.5 TOF ratio (i.e. very shallow blockade) [131]. One hundred and twelve patients were included (with fifteen lost to follow-up), i.e. twelve patients per dose of neostigmine. The primary endpoint was the time from injection of neostigmine or placebo to 1.0 TOF ratio. For cisatracurium, the time to reach 1.0 TOF ratio (minutes, median and range) that it would be pointless to reduce the neostigmine dose and that the best dose therefore remains 40 µg/kg. However, administering neostigmine at doses too high to reverse a very slight residual blockade is not exempt from side effects on neuromuscular transmission. In a study on sixty patients, two doses of neostigmine (20 and 40 µg/kg) were administered one, two, three or four hours after injecting a single dose of vecuronium [132]. In the forty patients receiving neostigmine 40 µg/kg, the TOF ratio had increased in thirty-two patients and decreased in eight of them. These patients had received neostigmine more than two hours after vecuronium and all had a TOF ratio > 0.9 prior to reversal. The decrease in TOF ratio

Rationale
The deeper degree of the neuromuscular blockade at the time of reversal, the longer the time between neostigmine administration and complete reversal (i.e. TOF ratio ≥ 0.9). This time ranges from 10 to 30 minutes [124]. After neostigmine administration it is therefore necessary to check the degree of recovery by monitoring the neuromuscular blockade.
R7.6 -It is recommended to adjust the dose of sugammadex according to ideal bodyweight and the intensity of neuromuscular blockade induced by rocuronium (Fig. 2). -when administered on reappearance of four responses to TOF stimulation, 1.0 mg/kg sugammadex reversed a rocuronium-induced neuromuscular blockade in less than five minutes. A dose of 0.5 mg/kg was also effective but slower (ten minutes) [134].
-when administered on reappearance of two TOF responses, a sugammadex dose of at least 2.0 mg/kg reversed a rocuronium-induced blockade in less than five minutes [135,136]. 4. very deep blockade: to reverse a very deep blockade (three and fifteen minutes after high doses of 1.0 or 1.2 mg/kg rocuronium) in less than five minutes, a sugammadex dose of at least 8.0 mg/kg was required [138][139][140].
An inadequate dose of sugammadex may be the cause of the phenomenon of recurarisation [141]. Thus, the sugammadex dose must be adjusted to the degree of blockade at the time of reversal. Monitoring remains crucial to adjust the sugammadex dose and should be continued after sugammadex administration to identify the potential occurrence of recurarisation [141].
The efficacy of sugammadex is decreased in elderly patients [142] and patients with severe renal failure (creatinine clearance < 30 mL/min) [143], especially in the case of deep blockade reversal (one to two responses to PTC). The rationale for using muscle relaxants in this setting is to prevent the motor consequences of generalised convulsions such as tongue biting, falls, dislocations and fractures. The specifications are for a fast-acting muscle relaxant that produces brief muscle relaxation during seizures. A systematic review recommends first-line use of suxamethonium. If there are contraindications, a short-acting non-depolarising muscle relaxant may be proposed [144]. A neuromuscular blockade is necessary during electroconvulsive therapy.
Suxamethonium remains the gold standard as a muscle relaxant in the vast majority of cases.

Rationale
Non-depolarising muscle relaxants and reversal agents are water-soluble drugs that are distributed in lean mass and extracellular volumes. As a result, apart from suxamethonium, their dose should be calculated based on lean bodyweight, which is higher in obese subjects than in subjects with a normal mass (Janmahasatian formula) [149].
No recommendation. There are insufficient data for any recommendations to be made concerning the interest of a deep blockade for laparoscopic surgery in obese patients.

Rationale
In laparoscopic surgery, compared to a strategy without a muscle relaxant, the neuromuscular blockade leads to improved visibility within the operating field, absence of movement and increased safety of the surgical procedure [97,108]. The depth of blockade remains an unresolved issue. The interest of a deep blockade defined by a PTC of one to two responses, compared to a moderate blockade or TOF stimulation involving two responses, has not been demonstrated as yet. Only one study reported a reduction in intra-abdominal pressure with better operating conditions in patients undergoing laparoscopic colectomy [104]. Another study reported marginally better operability, approaching significance in patients undergoing a cholecystectomy [98]. These studies involved mostly non-obese

Rationale
With steroid muscle relaxants, administration of neostigmine based on actual bodyweight at a dose of 0.04 mg/kg in combination with atropine 0.02 mg/kg is associated with a longer recovery time in obese subjects compared with non-obese subjects [150]. Sugammadex allows faster reversal than neostigmine [151]. Although the marketing authorisation states that the dose of sugammadex must be adapted to actual bodyweight, reversal of a partial blockade with two responses to TOF was obtained as quickly with a dose of 2.0 mg/kg calculated on the basis of ideal body mass (height in cm minus 110 in women and height in cm minus 100 in men) plus 40% (corrected weight) [152]. In the case of a deep blockade defined by the absence of response to TOF stimulation, the dosage should be increased to 4.0 mg/kg. In this case, adjusting the dosage to ideal bodyweight leads to reversal of the neuromuscular blockade in just over four minutes [153].
Children R8.6 -Other than situations for which rapid-sequence induction or the use of a depolarising muscle relaxant are indicated, the use of a non-depolarising muscle relaxant is probably recommended to improve intubation conditions during anaesthesia in children by intravenous induction.

Rationale
The SFAR Consensus Conference drawn up in 1999 did not recommend the use of muscle relaxants in children, whether induction is inhaled or intravenous. In fact, in France it is common not to use a muscle relaxant in children [154], and numerous hypnotic/opioid combinations have been reported [155]. However, in the context of intravenous induction, a meta-analysis of randomised studies in children reports improved intubation conditions when muscle relaxants are used [156][157][158][159][160][161][162][163]. In addition, the doses of opioid or hypnotics that allow tracheal intubation without a muscle relaxant are high and have haemodynamic effects that cannot be overlooked [159,161,164]. These results support those of a French cohort study [165]. In reference to the December 2017 ANSM alert on suxamethonium, depolarising muscle relaxants should not be used for intravenous induction, which is not part of a rapidsequence induction [7]. During inhaled induction, non-use of a muscle relaxant in children is a very common practice in France (92%) [154]. In this context, the duration of exposure to sevoflurane, its concentration, the agent(s) associated with the drug (opioid-propofol) affect the quality of intubation conditions but may also affect haemodynamic parameters [155,[166][167][168]. In all cases, achieving sufficient depth of anaesthesia and obtaining apnoea are the key conditions for success with this technique [169]. However, a muscle relaxant may be used during inhaled induction, especially in infants, for whom a prospective randomised study as well as a large-scale quality assurance study noted benefits associated with the inclusion of a muscle relaxant as regards intubation conditions and respiratory events [89,170]. These data (GRADE 1+) Strong agreement R8.8 -In conventional rapid-sequence induction, it is probably recommended that suxamethonium be given as a first-line drug for rapid-sequence induction in children. Where suxamethonium is contraindicated, use of rocuronium is probably recommended.

(GRADE 2+) Strong agreement
Rationale for 8.7 and 8.8 As in adults, it is recommended to limit the time between loss of consciousness and protection of the upper airways through intubation [172,173]. This time must be kept short since the duration of apnoea without hypoxemia is shorter in younger children [174]. Use of a muscle relaxant to improve intubation conditions, a muscle relaxant-free intubation technique or inhaled induction techniques are not recommended. Regarding fast-acting muscle relaxants, suxamethonium remains the expert choice. Age-appropriate doses of suxamethonium have been determined (< 1 month: 1.8 mg/kg, > 1 month and < 1 year: 2.0 mg/kg, > 1 year and < 10 years: 1.2 mg/kg, > 10 years: 1.0 mg/kg). Rocuronium at a dose higher than 0.9 mg/kg [175] may offer an alternative to suxamethonium, however, in 2018 sugammadex had not yet been granted marketing approval for children aged under two. The choice between suxamethonium and rocuronium will thus be based on the desired duration of neuromuscular blockade, the anticipated difficulties of intubation, and the presence and/or risk of ignoring underlying myopathy. The muscle relaxant most often compared to suxamethonium in the literature is rocuronium due to its rapid onset of action and the intubation conditions it offers [176]. The dosage is 0.6 to 0.9 mg/kg [175]. suxamethonium and non-depolarising muscle relaxants in terms of respiratory risk and difficult intubations [177]. In the latest Cochrane review, which assesses intubating conditions with rocuronium and suxamethonium, suxamethonium provides intubating conditions equivalent to or better than those obtained with rocuronium despite numerous biases in the various studies [8]. This review therefore concludes that suxamethonium should continue to be preferred (especially since its duration of action is shorter) and that rocuronium should only be used where suxamethonium is contraindicated [8,178]. Contraindications to suxamethonium include risk of malignant hyperthermia, muscle diseases with risk of rhabdomyolysis, hyperkalaemia, allergy and situations involving risk of hyperkalaemia [177].
Sugammadex has demonstrated its value in reversing the effects of rocuronium [179,180].
Indeed, a recent meta-analysis by Won et al. showed that sugammadex shortens mean time to achievement of ≥ 0.9 TOF ratio, with an extubation time comparable to neostigmine or placebo [179]. The conclusions of the studies regarding risk of anaphylaxis are contradictory.
The study by Reddy et al. shows a similar risk of anaphylaxis for suxamethonium and rocuronium [13]. Reitter et al. found a higher risk with suxamethonium in the results cited in the study appendix [181]. This risk may be lower with atracurium and cisatracurium, but it is present nonetheless [13,182]. Moreover, the risk of anaphylaxis with sugammadex does not appear to be negligible [183,184].
No recommendations. There is insufficient data on which to base any recommendations regarding the administration of a muscle relaxant in children to facilitate face mask ventilation, the insertion of supraglottic devices and management of related complications and surgical procedures and facilitate surgery or regarding the value of neuromuscular blockade monitoring for intraoperative tracheal intubation or the diagnosis and treatment of a residual neuromuscular blockade.

Argument
In paediatric anaesthesia, the frequency of residual neuromuscular blockade is estimated at 28% in children receiving a muscle relaxant and monitoring is therefore warranted [185].
There are no studies evaluating the occurrence of adverse events depending on whether a reversal agent is given. Similarly, there are no studies comparing the incidence of residual neuromuscular blockade based on age or the application of a neuromuscular blockade monitoring strategy. Suxamethonium is contraindicated in both these situations. In specific muscle disorders (myopathy, myotonia), it induces generalised contraction with rhabdomyolysis [186]. In the event of impairment of the nicotinic acetylcholine receptors (nAChRs) at the motor end plate through up-regulation (chronic damage of motoneurons, extensive and deep burns, prolonged critical illness), suxamethonium can cause life-threatening hyperkalaemia a few days after constitution of the neurological and/or muscle lesions, and over an extended period of time [186][187][188].
R8.10 -Monitoring of neuromuscular blockade is probably recommended following muscle relaxant use in patients with neuromuscular disease.

Rationale
In cases of down-regulation of nAchRs (myasthenia), muscle relaxants are not contraindicated but significant pharmacodynamic changes have been reported. With suxamethonium, resistance is observed (decreased potency and need to increase the dose to achieve the same effect) [189,190]. For non-depolarising muscle relaxants, there is an increase in sensitivity and in duration of action, with reduced intraoperative dose requirement. A 50-75% reduction in the recommended dose is common with atracurium and cisatracurium [191,192]. This reduction in dose requirement correlates with the severity of myasthenia. Neuromuscular blockade monitoring is recommended to avoid overdosing.
Evaluation in myasthenic patients of the TOF ratio by EMG in the hypothenar hand muscles prior to muscle relaxant administration predicts sensitivity to non-depolarising muscle relaxants. If the TOF ratio is less than 0.9 before the neuromuscular blockade, sensitivity to muscle relaxants is greater and the injected doses must be lower than in myasthenic subjects with a ratio greater than 0.9 [193]. In the case of primary muscle damage, there is a very significant increase in sensitivity to rocuronium (reduced dose requirement). A comparative study showed that after a dose of 0.6 mg/kg of rocuronium, onset and recovery times were significantly longer for Duchenne muscular dystrophy patients compared to controls [194]. In cases of up-regulation of nAchRs, sensitivity to non-depolarising muscle relaxants is reduced and higher doses are required to achieve the same effect [195]. The small populations in the various studies warrant GRADE 2 classification.
R8.11 -Administration of sugammadex is probably recommended for reversal of a residual neuromuscular blockade following the use of a steroidal muscle relaxant in patients with neuromuscular disease.

Rationale
The risks of residual neuromuscular blockade are increased in neuromuscular disease. In myasthenia patients, neostigmine may interfere with long-term treatment. In primary muscle damage, neostigmine and atropine are difficult to deal with due to drying of secretions (atropine), potential rhythm and conduction disorders (both), central effects (atropine), slow response and effects on muscle action potential (neostigmine). Case series have described the use of sugammadex as a reversal agent for rocuronium in patients with neuromuscular disease [196,197]. The results are comparable with those observed in clinical trials conducted in subjects without neuromuscular disease (see R-7.2) regarding efficacy and time to onset of action. The small sample sizes of patients in the different studies warrant GRADE 2 classification.

Renal/hepatic failure, elderly subjects
R8.12 -The use of a benzylisoquinoline muscle relaxant (atracurium/cisatracurium) is probably recommended in cases of renal/hepatic failure.

Rationale
In this setting, the pharmacodynamics of muscle relaxants excreted by kidney or liver are modified and their elimination is delayed. This may result in extended duration of action of these agents, particularly with repeated injections or continuous infusion [198]. In addition, interindividual response variability is greater in these populations, resulting in increased difficulty in dealing with these agents. Rocuronium is mainly eliminated in urine and bile [199] and its clearance is thus reduced in renal failure patients [200] and cirrhotic patients, in whom there is wide variability in the duration of action of repeated injections. [201]. The pharmacokinetics and pharmacodynamics of atracurium, roughly half of which is eliminated by organ-independent reactions (Hofmann reaction and ester hydrolysis) and half by metabolism or excretion [202], are similar in subjects with and without kidney and liver failure [203,204]. On the other hand, its active metabolite, laudanosine, accumulates in patients in renal failure but does not reach concentrations causing adverse effects, even after infusion for up to 72 hours [205]. Cisatracurium, one of the ten isomers of atracurium, elimination of which is overwhelmingly non-enzymatic [206], also has similar pharmacokinetic and pharmacodynamic profiles in patients with and without renal and hepatic failure [207]. Since this agent is more potent than atracurium, the doses, and hence the amounts of laudanosine generated, are significantly lower.
R8.13 -It is recommended not to modify the initial dose in renal/hepatic failure patients, irrespective of the type of muscle relaxant used.

(GRADE 1+) Strong agreement
Rationale patients and in the elderly, the time to onset of action remains unchanged [208]. In cirrhotic patients, the efficacy of the initial dose of rocuronium is reduced since the increase in distribution volume results in a lower concentration [201]. However, this effect is particularly remarkable at low doses and is no longer clinically noticeable at intubation doses [201]. The same is true of atracurium [209] and cisatracurium [210] in renal/hepatic failure.
Consequently, to ensure effective concentrations of muscle relaxants during intubation, it is necessary and sufficient to administer the usual dose in both renal and hepatic failure patients and cirrhotic patients.
R8.14 -When using sugammadex in cases of renal failure, it is probably recommended to administer it at the usual dose.

Rationale
Sugammadex is eliminated in urine and thus accumulates in renal failure patients [200].
However, clinical studies have shown that the efficacy of sugammadex is maintained in renal failure patients at the same doses as in subjects with normal renal function, without any signs of recurarisation [211]. Sugammadex may be removed by dialysis, as may sugammadexrocuronium complex [212].
No recommendation. There is insufficient data in the literature to establish a recommendation concerning the use of muscle relaxants in the elderly.