Succinylcholine versus rocuronium for rapid sequence intubation in intensive care: a prospective, randomized controlled trial

Endotracheal intubation in critically ill patients is a high risk procedure containing the danger of hypoxia and cardiovascular collapse. The method of choice for emergency intubation in the intensive care unit (ICU) is a rapid sequence induction (RSI). Because of its fast onset, succinylcholine is the most commonly used neuromuscular blocking drug in RSI. Due to its depolarizing mechanism of action resulting in an increase in extracellular potassium, succinylcholine is contraindicated in a number of circumstances and diseases frequently present in critically ill patients [1]. Rocuronium has the most rapid onset of the currently available non-depolarizing neuromuscular blocking drugs. As the only contraindication to rocuronium is the very rare occasion of allergy, this agent is regarded as an attractive alternative to succinylcholine [2, 3]. Assessing current clinical practice by analyzing large contemporary trials reveals a wide variation in the use of rocuronium or succinylcholine for RSI in the ICU [4‐8]. However, neither agent has been so far tested against each other or any other neuromuscular blocking agent in an adequately powered randomized trial in critically ill patients.

Compared to succinylcholine, rocuronium is associated with less optimal intubation conditions [3, 9] and a longer intubation sequence [9] in the operating theatre. As the most frequent complication of RSI in the ICU is severe hypoxemia [4, 7] the combination of less optimal intubation conditions and longer intubation sequence may be of relevance. Given the low rate of hypoxemia during RSI in the operating theatre, available studies are underpowered for this outcome. The aim of this prospective randomized controlled trial was, therefore, to compare the incidence of hypoxemia after rocuronium or succinylcholine in critically ill patients requiring an emergent RSI.

The study started in August 2006 and ended with the inclusion of the 420^th patient in June 2010. Figure 1 shows the patients' flow. Of the 401 intubations 333 (83%) were accomplished on the first, 57 (14.2%) on the second, 9 (2.2%) on the third, and 2 (0.5%) on the fourth attempt. No periprocedural death occurred. Table 2 shows the indications for the intubations performed. Demographics of the patients included in the study and the registry are shown in Table 3.

During the course of an emergent intubation the incidence and severity of oxygen desaturations did not differ between critically ill patients undergoing RSI with succinylcholine and those undergoing RSI with rocuronium as the neuromuscular blocking agent. The mean intubation sequence was 14 sec shorter after succinylcholine than after rocuronium. Succinylcholine and rocuronium resulted in similar intubation conditions and a similar incidence of intubation related complications.

A recent Cochrane Review demonstrated that for a RSI succinylcholine created better intubation conditions than rocuronium [3]. However, all patients included in the Cochrane Review were intubated in the operating theatre. Conditions in intensive care and in the emergency department may differ in several important aspects from those in the operating theatre (for example, severity of patients' illness, limited possibility to perform a pre-intubation airway assessment, less ideal ergonomic conditions) so that findings from one setting are not necessarily applicable to the other. Particularly, the incidence of relevant intubation-related complications is very small in the operating theatre so that available studies are underpowered for these outcomes. In keeping with previous work, the present study demonstrates that emergent intubations in intensive care are associated with a high rate of immediate and potentially life-threatening complications like oxygen desaturations and failed intubation attempts [4, 5, 15].

In contrast to the majority of studies performed in the operating theatre [3], the present study demonstrated no difference in intubation conditions between succinylcholine and rocuronium. In emergency department RSI, Lauren et al. reported that succinylcholine resulted in less body movements as a reaction to intubation than rocuronium, while there was no difference between the two drugs with regard to the degree of vocal cord movements [16]. In emergent RSI in the operating theatre, Sluga et al. reported that the difference in the overall intubation scores between succinylcholine and rocuronium was entirely due to a difference in the sub-score assessing the intubation response, that is, limb movements and/or coughing, while there was no difference in the remaining sub-scores assessing the ease of laryngoscopy or the conditions of the vocal cords [9]. The previous finding of similar sub-scores for laryngoscopy and vocal cords is confirmed by our results. As limb movements and/or coughing were only rarely observed in the present study, sub-scores for the intubation response were very high in both groups and did not result in a difference in the overall score for intubation conditions. This absent response to intubation in the majority of critically ill patients is most likely due to the severity of the underlying illness. Based on the present results and previous findings [9, 16] we propose that the difference reported for intubation conditions between succinylcholine and rocuronium [3] results entirely from a difference in the response to intubation, an event occurring after the completion of the intubation with marginal relevance for patients' safety. By contrast, succinylcholine and rocuronium do not differ in two aspects of intubation conditions highly relevant for patients' safety - the ease of laryngoscopy and the conditions of the vocal cords.

During emergent intubations in the operating theatre, succinylcholine allowed for a 35 sec earlier completion of the intubation sequence than rocuronium [9]. In the present study involving critically ill patients, this favorable effect of succinylcholine was reduced to 14 sec, that is, to approximately 15% of the total length of the intubation sequence with rocuronium. In almost 40% of our patients visible fasciculations had not occurred or had not ceased within 45 sec after the injection of succinylcholine, and, according to study protocol, laryngoscopy was started in these patients at the same time as in the rocuronium group. To the best of our knowledge, a delayed onset of succinylcholine-induced fasciculations in critically ill patients has not been reported so far. As severe illness may have a profound effect on the neuromuscular system [17] it is tempting to speculate that the severity of underlying illness is responsible for the difference in the onset of fasciculations and, hence, the reduced advantage of succinylcholine with regard to rapid completion of the intubation sequence between patients in the operating theatre and patients in intensive care.

A limitation of this trial is the lack of a double-blind design. However, masking the effects of drugs like succinylcholine that have visible effects (fasciculations) is inherently difficult. A RSI in critically ill patients is a high-risk procedure requiring the full attention of an appropriately trained physician. Since in our settings the simultaneous achievement of perfect blinding and optimal patient safety was not feasible, we opted for a single-blind study design. Using experienced operators and a checklist (containing all relevant aspects recently found to decrease complications related to intubation in intensive care [5]), every effort was made to protect the patients' safety. Thus, our results are not necessarily generalisible to settings with less stringent procedural guidelines and/or less experienced operators. Our findings were obtained with an induction sequence of fentanyl and propofol or etomidate and cannot be extrapolated to other drugs and/or doses.

The dose of rocuronium was chosen based on the best evidence available at the time of the design of the trial, that is, the Cochrane Review of 2003 [2]: For the outcome acceptable versus suboptimal intubating conditions, this review revealed no significant difference between succinylcholine and rocuronium at a dose of 0.6 to 0.7 mg/kg. Moreover, a sensitivity analysis within the propofol induction group (n = 640) for dose of rocuronium used (0.6 to 0.7 mg/kg versus 0.9 to 1.0 mg/kg) demonstrated that dose did not alter intubating conditions. Weighing the risk of prolonged neuromuscular blockade associated with higher doses of rocuronium against the lack of convincing evidence of better intubating conditions, we chose the smallest dose proven to be effective, that is, 0.6 mg/kg. However, the 2008 update of the Cochrane Review demonstrated an advantage of succinylcholine against lower (0.6 to 0.7 mg/kg), but not against higher (> 0.9 mg/kg), doses of rocuronium for the outcome acceptable versus suboptimal intubating conditions [3]. As the difference is small (risk ratio 0.95, 95% confidence intervals 0.90 to 0.99) and, as extensively discussed above, appeared to be mainly caused by the response to intubation [9], we decided not to modify the protocol of our ongoing trial. Our trial demonstrates identical intubating conditions with 1 mg/kg succinylcholine and 0.6 mg/kg rocuronium. Thus, in critically ill patients there is not an advantage in choosing higher doses of rocuronium than 0.6 mg/kg.

The adverse effect profile of succinylcholine is of concern and some authors even considered its use in the ICU as obsolete [1]. Optimal intubation conditions [2, 3], a short intubation sequence [9], and the return of spontaneous respiratory activity within 5 to 10 minutes are the main arguments of the proponents of the use of succinylcholine in the ICU [18, 19]. The present study is the first formal comparison of succinylcholine and rocuronium in an adequately powered randomized controlled trial in critically ill patients. Our results demonstrate that in emergent intubations in intensive care 1) the incidence and severity of hypoxaemia does not differ between rocuronium and succinylcholine; 2) intubating conditions do not differ between rocuronium and succinylcholine; 3) the advantage of succinylcholine with regard to the length of the intubation sequence is markedly reduced to a difference of questionable clinical significance; and 4) the incidence of failed intubation attempts does not differ between rocuronium and succinylcholine. Lee et al. reported that the reversal of profound rocuronium-induced neuromuscular block with sugammadex was significantly faster than spontaneous recovery from succinylcholine [20]. Taking together its favourable safety profile, the availability of a rapid-acting antagonist, and our present results, rocuronium appears to be more than a suitable alternative to succinylcholine for emergent RSI intubations in acutely ill patients.

■ This is the first formal comparison of succinylcholine and rocuronium for use in emergent rapid sequence induction intubation in intensive care in an adequately powered randomized controlled trial.

■ In critically ill patients, undergoing an emergent rapid sequence induction intubation, the incidence and severity of oxygen desaturations, the quality of intubation conditions, and the incidence of failed intubation attempts did not differ between succinylcholine and rocuronium.