Which port in a storm? Use of suxamethonium without intravenous access for severe laryngospasm

The use of intramuscular suxamethonium has gained pole position in paediatric anaesthetic practice as the treatment for severe laryngospasm during an inhalational induction prior to intravenous access being achieved. Despite reservations due to long onset time, there are reports confirming its efficacy and letters strongly supporting its use in clinical practice [1, 2]. Donati and Guay, on the other hand, in an editorial in 2001 [3], entitled ‘No substitute for the intravenous route’, argue against the use of intramuscular suxamethonium for treatment of laryngospasm on the basis of long onset time. He states that ‘all possible strategies to relieve laryngospasm should be considered’, and that ‘help should be sought to search for intravenous access, and no site, including femoral, should be rejected.’ He does not, however, offer a solution should intravenous access fail in the limited time available. So, if the infant, child or, indeed, adult develops laryngospasm and is desaturating during a gaseous induction before intravenous access is obtained, is there a substitute for the intravenous route to administer suxamethonium? Suxamethonium can indeed be administered by the intramuscular route, the intralingual route, and the intra-osseous route. Which port in a storm? The first reports concerning the effects of intramuscular suxamethonium date from the 1950s [4, 5]. McDonald and Bryce-Smith [4] evaluated the use of intramuscular suxamethonium on the premise that, if intramuscular curare worked for the South American Indian on the tip of an arrow, then rapidly acting suxamethonium could work for the anaesthetist struggling to find veins. They found predictable results in doses of 1 mg.lb−1 (2 mg.kg−1) when mixed with hyaluronidase, and 2 mg.lb−1 (4 mg.kg−1), without hyaluronidase. They looked at adults and children and found the time to apnoea to be 2–3 min. Liu et al. [6], in 1981, evaluated suxamethonium in doses of 2 mg.kg−1, 3 mg.kg−1 and 4 mg.kg−1 by the intramuscular route in children. The maximum twitch depression at the adductor pollicis muscle varied widely after intramuscular administration of 2 mg.kg−1 and 3 mg.kg−1. Some in the 2 mg.kg−1 and 3 mg.kg−1 groups achieved a maximal twitch depression of only 10–30%. In those given 4 mg.kg−1, the results were more predictable. All patients achieved a maximal twitch depression of greater than 80%, and all but one was greater than 90%. The duration of action for the 3 mg.kg−1 and 4 mg.kg−1 dose was similar, in the region of 21 min, showing characteristics of a dual block. Two simultaneous injections to increase the surface area for absorption made no difference to onset or duration. Work looking at speeding the onset of intramuscular suxamethonium has been generally disappointing. The use of hyaluronidase has thrown up conflicting results [7]; however, there is animal work showing the ‘time to collapse’ in goats to be shorter in the hyaluronidase group by 25%[8]. Using different concentrations of suxamethonium (2% or 5% solutions) made no difference [9]. So, is the onset fast enough for the treatment of severe laryngospasm? Liu and colleagues had doubts [6]. They stated that, ‘In an already asphyxiated child, the 3–4 min required for maximal twitch depression after 4 mg.kg−1 argues against its use.’ They do go on to say, however, ‘Nevertheless, clinical experience with intramuscular suxamethonium in children has been that airway control after laryngospasm is achieved in less time than the time to maximal twitch depression’. All of these papers were assessing intramuscular suxamethonium for intubating conditions. In the management of laryngospasm, oxygenation is the initial goal and, therefore, full relaxation is probably not necessary. So, does laryngospasm break before the time to maximal twitch depression? The answer may lie in one of the first reports of the use of intramuscular suxamethonium by Beldavs in 1959 [5]. In this report, the author gave suxamethonium to 40 conscious children intramuscularly before the introduction of the anaesthetic. He noted that ‘if the patient was crying prior to induction, he lost his voice in 30–45 s after the injection, although the muscles of the extremities still seemed to be very strong’. Beldavs' study suggests weakening of the laryngeal musculature before onset of peripheral skeletal muscle paralysis. This interesting observation was later followed by some elegant work by Donati et al. [10] showing exactly that. The onset of neuromuscular block was indeed faster in centrally located muscles, such as the diaphragm and the larynx. It follows that oxygenation should be possible before the adductor pollicis muscle block reaches maximum. This observation could be supported further by an interesting case report in which very small intravenous doses of suxamethonium were used to break laryngospasm during bronchoscopy in adults [11]. The dose used was 0.1 mg.kg−1 and at this dosage spontaneous respiration was maintained while laryngospasm was broken. What are the downsides to intramuscular suxamethonium? Hannallah et al. looked for arrhythmias with intramuscular suxamethonium with and without atropine in small children [12]. No serious disturbances of rhythm were detected in any patient, nor are there any reports of cardiac arrhythmias in other papers. One question mark relates to the development of pulmonary oedema [13]. There is one report of three infants given intramuscular suxamethonium for airway control routinely prior to intubation. There is no indication of airway obstruction or indeed, of laryngospasm. In these infants, frank pulmonary oedema developed within 4–8 min following the administration of intramuscular suxamethonium. The mechanism of this development was unexplained. So, intramuscular suxamethonium seems overall to be a winner. Practically, the intramuscular route is accessible. One can administer suxamethonium into the deltoid muscle without leaving the airway. One must, of course, have sufficient drug prepared beforehand and this may mean more than the conventional 100 mg to give the recommended dose of 4 mg.kg−1. It should work within 1 min, but there remains some doubt. There is the question of absorption in a peri-arrest situation where muscle perfusion is poor. Some studies also report that roughly 10% of patients did not attain more than 30% maximal twitch reduction. This could be due to subcutaneous injections and it remains a risk, but even this is probably enough to break laryngospasm. The tongue acts as an extremely vascular muscle which retains blood flow to a greater extent compared to peripheral skeletal muscle in times of reduced perfusion. There is one report only describing intralingual suxamethonium for severe laryngospasm in an 80-kg 65-year-old lady via the intra-oral, intralingual route [14]. Tight laryngospasm and a firmly closed jaw developed following induction. The intravenous cannula was lost by a ‘helpful medical student’ and attempts to reinsert it quickly failed. There was a gap in the incisors just large enough to insert a needle and the injection was made into the tongue with 100 mg of suxamethonium. Within 30 s the jaw relaxed and she was intubated. The characteristics of intralingual suxamethonium were studied by Mazze and Dunbar in 1968 [15]. It was postulated to have a quicker onset of action than peripheral intramuscular suxamethonium. A dose of 0.5 mg.lb−1 (1.1 mg.kg−1) was administered to 60 children via the intra-oral, intralingual route (using a 25-gauge needle into either surface of the tongue). The mean onset time to apnoea was 75 s, compared to 35 s for intravenous administration and 210 s for the intramuscular route. They observed that apnoea was distinct in the intravenous and intralingual groups but was less distinct in the intramuscular group. Respiration returned at 7 min 15 s in the intralingual group. This compared with 5 min in the intravenous group and 9 min for intramuscular group. However, they reported a high incidence (> 50%) of arrhythmias (bigeminy, premature ventricular contractions and sinus bradycardias) in patients given intralingual suxamethonium. One patient with bigeminy developed ventricular tachycardia. Most arrhythmias, however, were of the ‘vagal type’, and all were self-correcting and required no treatment. The other groups suffered no arrhythmias. The high number of arrhythmias with intralingual suxamethonium could not be explained at that time. An alternative to an intralingual injection is the submental, extra-oral approach. This may be used when clenched jaws preclude access to the mouth. Redden et al. [16] compared this approach to the intramuscular route. They used 3 mg.kg−1 in the intramuscular group and the same dose intralingually. They found a mean onset time of 265 s following submental injection, compared to 295 s with intramuscular suxamethonium. They looked additionally at the effect of digital massage with the extra-oral, submental approach. The mean onset time decreased to 133 s. This group was alone in achieving 100% maximal twitch depression. There were no complications, including intra-oral bleeding. The intra-oral, intralingual route is a better bet to gain maximal relaxation within 60 s, but there is relatively little work reported on its use. There remain the question of associated arrhythmias and the practical issue of administering the drug (removing tight continuous positive airway pressure in a hypoxic patient to inject). One could argue that only muscle relaxation will aid oxygenation in the deteriorating situation and thus it is reasonable to do this. A dose of 3 mg.kg−1 is recommended, as with the extra-oral, submental, intralingual injection. The latter has the advantage, like the intramuscular route, that one can keep the facemask tightly applied, but the onset is somewhat longer even with digital massage. The use of intra-osseous infusions in recent years has gained acceptance for rapid access to the circulation in children up to 6–8 years of age [17]. Tobias and Nichols reported two children brought to the emergency department requiring intubation [18]. Both children had low Glasgow Coma Scores and clenched teeth. Intravenous access proved impossible and an intra-osseous needle was inserted. Induction of anaesthesia then followed with thiopental and intra-osseous suxamethonium. Relaxation occurred within 30–45 s. Animal studies [19] have verified the similarity between intravenous and intra-osseous suxamethonium. Six sheep were anaesthetised on three separate occasions, each time receiving intra-osseous, intravenous or intramuscular suxamethonium. The time to respiratory arrest and 100% loss of the forefoot twitch was recorded. The average time from administration to respiratory arrest was 30.8 s (intravenous), 230 s (intramuscular), and 57.5 s s (intra-osseous). Intra-osseous access was achieved within 30 s. Choosing this route means that intra-osseous needles need to be quickly available and that someone is able to use them. A ‘friendly surgeon’ could be utilised, should you be that lucky. Kruse et al. claim that intra-osseous access can be obtained within 60 s [20], but to date there are no reports of its use to treat severe laryngospasm and the speed test under pressure has not been done. Leaving a patient in laryngospasm is potentially serious; post obstructive pulmonary oedema and/or hypoxic damage are both possible outcomes [21]. In an Australian study looking at reported incidents of laryngospasm, there was one cardiac arrest and 35% suffered major physiological changes; 4% of patients were thought to have suffered post obstructive pulmonary oedema [22]. It is therefore essential to treat laryngospasm promptly. There are, indeed, substitutes for the intravenous route when faced with severe laryngospasm. Sufficient muscle relaxation and oxygenation can be achieved within 60 s by administering suxamethonium by any of the routes described above. The route chosen will depend very much on individual circumstances and preference, but a rapid decision to give suxamethonium by the intramuscular, intralingual or intra-osseous route could save the day.