93 Continuing Medical Education Article Laryngospasm in

W. L. Roy MD FRCPC, J. Lerrrmn MD FRCPC paediatric anaesthesia

Laryngospasm occurs frequently during the administra- inspiratory pressure gradient increases. Fink differentiated tion of anaesthesia to paediatric patients. ~ The reported stridor from laryngospasm - the former being an intermit- incidence of laryngospasm in patients aged 0-9 years is tent closure of the glottis alone and the latter, a prolonged 17.4/1,000 patients. ~ This incidence is almost twice that interruption of respiration due to a ball valve mechanism of the total population studied (8.7/1,000 patients). of the vocal cords, false cords and supra-giottic tissue. Although most instances oflaryngospasm resolve without Fink noted that "when laryngeal spasm sets in, the serious complications, complete laryngeal obstruction expiratory effort becomes prolonged and closure of the can result in intense laryngospasm, gastric aspiration, larynx and contraction of the abdominal wall become cardiac arrest and death. ~Because of the high incidence of continuous." Laryngospasm in fact may be an extreme laryngospasm and the severity of potential complications, form of cough. a review of laryngospasm in children, its etiology, Clinicians commonly observe prolonged glottic closure treatment and prevention, is presented. in response to glottie or supra-glottic mucnsal stimula- lion. Laryngeal closure cominucs long after cessation of Pathophysiology mueosal stimulation. Suzuki believed that laryngospasm The pathophysiological mechanism of laryngospasm dur- was an exaggerated laryngeal response to stimulation of ing anaesthesia is unclear.: Most practising anaesthetists the superior laryngeal nerve. 4 Although his studies believe that laryngospasm is precipitated by closure of the involved animals, he made some interesting observations true vocal cords. This closure results in either complete or which might very well have clinical relevance to humans. incomplete airway obstruction. However, Keating3 ques- He noted that synaptic transmission in the superior tioned the mechanism nf laryngospasm. He stated that laryngeal nerve was reduced during deep barbiturate either the true vocal cords alone or both the true and false anaesthesia. This contradicted the belief that barbiturates cords become apposed in the midline and close the glottic actually aggravated laryngospasm. He also observed that inlet during laryngospasm. The role of the false vocal hypoxia abolished post-synoptic potentials and led to an cords in laryngospasm is unclear. overall depressive effect on all reflex neural activity in the Debate over the mechanism of laD,ngospasm continues animal model. This latter observation supported the notion from both anatomical and neurophysiological points of that severe hypoxia n'tay actually tenlfi.nate laryngospasm view. We found it most helpful to return to Fink's early It appears that the results of additional animal studies ex- review of laryngospasm: for a clearer description of this plain in part our clinical observations.5.6 Laryngospasm is entity. Laryngeal closure involves apposition of struc- a sustained event, which can be abolished by deep tures at three levels: (1) the supraglottic fulds, (2) the false anaesthesia, severe hypoxia and hyperearbia. vocal cords, and (3) the true vocal cords. With the aid of lateral x-rays of the neck, tomography and electromyog- Etiology raphy, Fink found that the human larynx was more than Several investigators have studied laryngospasm during just a shutter. He proposed a dual mechanism for closure anaesthesia. ~,70lsson et al., ~ during an 11 year prospec- ef the tarynx: first, a shutter effect (the vocal cords) and tive study, found an overall incidence of 7.9tl,000 secondly, a ball valve effect (the false cords and the anaesthetics or 8 ~7/1,000 patients. The incidence in chil- redundant supra-glottic tissue). The ball valve effect dren 0-9 years of age was higher, 17.4/1,000 patients, depends on the shortening of the thyrohyoid muscle, an and within this age group, infants 1-3 months of age had extrinsic muscle (Figure l). The soft tissues of the the greatest incidence (more than three times the rate in supra-glottic region then become rounded and redundant any other age group). Other factors associated with an and are drawn into the laryngeal inlet as the translaryngeal increased incidence of laryngospasm in children were:

CAN J ANAESTH 1988 / 35: I ? pp93-8 94 CANADIAN JOURNAL OF ANAESTHKfilA

FIGURE | Sectional view of larynx (adapted from Fink 8R. The Etiology and Treatment of Luryngeal Spasm. Artcsthe~iolagy 1956; 17; 569-77, with pet-m~s~i~n),

extubation (42/1,000), presence of a nasogasttic tube However, there are no data to support such a notion. Com- (48.5/1,000) and oral endoscopy and esophagoseopy pared to halothane and enflurane, isoflurane has been (48.5/1,000). The incidence of laryngospasm was highest associated with a significantly greater incidence of laryn- (95.8/1,000) in children with respiratory tract infections. gospasm during induction of anaesthesia in children, t4 Barbiturates have been identified as a group of drugs Light anaesthesia is another important cause oflaryngo- which predispose to the occurrence of laryngospasm. spasm. In 1939, Cole addressed the problem of respiratory Olsson and Hallen ~ found a slightly greater incidence of obstruction during thyroidectomy and concluded that '*if a laryngospasm in children who were anaesthetized with a patient is anaesthetized deeply before the operation is barbiturate compared to those anaesthetized without a begun, laryngeal spasm wil~ be much less likely to barbiturate. Some authors have suggested that laryngeal develop. ''ns Reflex closure of the larynx was thought to be spasm is a direct complication of the use of thiopentone, s- i 0 a response to stimulation of visceral nerve endings in the They have speculated that laryngospasm is due to para- pelvis, abdomen, thorax, or larynx itself; and when such sympathetic overactivity in the airway inlet resulting from stimulation led to respiratory obstruction this was most a direct action of thiopentone. Others have proposed that often the result of indequate anaesthesia.2 Because many parasympathetic overactivity increases as the dose of the patients are intubated during anaesthesia, the risk of barbiturate increasesJ ~ Animal studies have been contra- laryngospasm due to light anaesthesia during surgery is dictory. In 1935, an examination of the effect of short minimal. However, /aryngospasm may occur following acting barbiturates on the glottic opening in eats suggested extubation during light anaesthesia in children.16 Sadly. a relationship between barbiturates and parasympathetic- our technology has not eliminated this serious eomplica, induced laryngospasmJ2 A more recent study in 1977 tion b~t has simply deferred it from the intraopera'five found that barbiturates depressed the reflex closure of the period to the postoperative period. glotti~ in cats.4 The debate concerning the effect of short- acting barbiturates on the laryngeal reflex mechanism continues unresolved. Management Volatile anaesthetic agents have also been associated Incomplete airway obstruction is generally associated with laryngospasm. 13.14 In the eat, laryngospasm can be with an audible inspiratery or expiratory sound, which is triggered by instilling ether ot halothane directly into the best heard with a precordial stethoscope. If the obstruc- trachea: an effect not seen with methoxyfiurane. L3 It was tion becomes progressively worse, tracheal tug, and suggested that laryngospasm might be avoided if the con- paradoxical respiratory movements of the thorax and centrations of these volatile anaesthetics were restricted. abdomen will develop. Should the obstruction progress Roy and Larman: LARYNGOSPASMIN PAEDIATRICS 95

I INCOMPLETEAIRWAY 1 OBSTRUCTION L i PRESSUREAPPLYGENTLE WITH 100% POSITIVE OXYGEN 1 ,~ ~ f DEEPEN ANAESTHESIA > | ~ = I WITH INTRAVENOUSTHIOPENTONE/ =~1 ...... ~ Sr~s.,ZETHE A, FtWAY ANO /

INCREASE OR" DECREASE CONC, NO II.qPROVENEN't OF VOLATILE AGENT

r,v SO =. 'OHO.,NE A;.OP,NE1 o l ~llb] VENTILATEWITH 100% OXYGEN / =4""'""' i NO'NTU=ATE =TH E'T"TO~ J

I STABILIZEANAESTHETIC PATIENT AND RESUME I

FIGURE 2 Algorithm displayias eourzc el management for incomplete airway obstruction,

from incomplete to complete, the audible sounds cease, either side of the larynx and presses the aryepiglottic folds leaving only visible signs of airway obstruction. more firmly against each other" (Figure 1). This manoeuvre Incomplete upper airway obstruction can often be is commonly taught as an effective means of managing relieved with one of several therapeutic manoeuvres laryngeal spasm. However, in the presence of complete (Figure 2). The first is to remove the irritant stimulus,2 by obstruction, oxygen will fill the stomach instead of the terminating surgical stimulation of visceral nerve endings lungs, in the presence of complete obstruction of the air- or by removing debris from the larynx. The second is to way, the temporomandibular joint should be dislocated deepen the level of anaesthesia. The third is to facilitate anteriorly by applying pressure to the ascending rami of ventilation by applying gentle continuous positive airway the jaw. z This manoeuvre lengthens the thy~hyoid muscle pressure with 100 per cent oxygen via a tight-fitting face and unfolds the soft supraglottic tissue. If dislocation of mask. the tempero-mandibular joint fails, then atropine and Complete airway obstruction shares many of the succinylcholine should be administered intravenously. clinical signs with incomplete obstruction - tracheal tug, However, laryngospasm may occur when intravenous in&awing of the chest wall and marked abdominal access is not available. In this instance, intramuscular respiration. However, the absence of respiratory sounds is succinyleholine (4 mg.kg-I) is recommended, t7 After the "sine qua non" of complete laryngospasm. Continuous imramuscular succinylcholme, the vocal cords will usually monitoring of respiratory sounds with a precordial stetho- relax sufficiently within one minute to permit ventilation, scope will lead to an early diagnosis of airway obstruction and within several minutes to facilitate tracheal intuba- and should result in early therapeutic intervention. A tion. Pulmonary oedenm has been reported following the precordial stethoscope should be a routine monitor during administration of intramuscular succinylcholineJ s The the administration of anaesthesia to paediatric patients. mechanism of this is unclear, lntralingual atropine and When complete obstruction of the upper airway occurs, succinylcholine are not recommended in children anaes- what is the best management? Fink 2 wrote: "Such a spasm thetized with halothane/aitrous oxide/oxygen ~9 since cannot be broken by bag pressure through a mask. Forced vcntricular arrhythmias have been reported. If the laryngo- inflation of the pharynx distends the pyriform lbssas on spasm is sustained and the child becomes hypoxic, it may 96 CANADIAN JOURNAL OF ANAESTHE$[A

I START CPR . CONSIDER 1 CRICOTHYROTOMY OR TRACHEOTOMY

rCOMPLETEAIRWAY OBSTRUCTION1 FAILEDl INTUBATJON DUE TO LARYNGOSPASM J

REPEAT LARYNGOSCOPY |SPRAY CORDS WITH LIDOCAINE| l ATTEMPT,NTUBATION l L CALL FOR HELP ) I APPLY FIRM UPWARD PRESSURE1 BEHIND ANGLE OF MANDIBLE FAILEDl INTUBATION ADMINISTER 100% OXYGEN BY MASK I IMMEDIATE LARYNGOSCOPY INTUBATE WJTHOUT RELAXANTS~ VENTILATE WITH 100% OXYGEN

I LISTEN AND WATCH FOR ANY OXYGEN ENTRY

~UCCIN~LeHOLmE ,.S ~,. / ~g. tV.) ATROPINE 0.02 rag. / kg. I.V. IF | WITH I,V, ACC ES~, BRADYCARDIA IS PRESENT | v'

APPLY GENTLE POSITIVE i PRESSURE TO THE AIRWAY 1 I VENTILATE WITH 100% OXYGEN INTUBATE WITH AN APPROPRIATE SIZED ENDOTRACHEAL TUBE

CONSIDEREITHER SUB . LINGUALOR INTRAMUSCULARADMINISTRATION OF SUCCINYLCHOLINEAND A]ROPINE. :SEETEXT FOR FURTHERDETAILS

FIGURE 3 Algorithm displaying course of managerneta foz compLete ;tit'way cbstruclion.

bc necessary to intubate without muscle relaxation. 2~This Prevention may be preferable to waiting for the effects of succinyl- Prevention, of course, is the best therapy for [aryngo- choline in a child who has Iaryngospasm and who is spasm. Olsson et al. I stated that "to minimize the risk of bradycardic. Under these extreme conditions the vocal perioperative laryngospasm, it is necessary to be aware of folds may be sprayed directly with lidocaine (Figure 3), in the factors which increase the risk of laryngospasm." The order to relax the larynx and facilitate intubation. The high risk factors Olsson enumerated included pre-existing mechanism of this relaxation is unknown hut may be due respiratory problems, a history of previous anaesthetic to the d[zect effect of local analgesia or to a prolonged complications, surgery including endoscopy and hypo- effect of hypoxia. If after all of these measures the airway spadias repair, and those with a nasogastric tube in situ. has not been secured, then cricothyrotomy or emergency Intravenous lidocaine 2 mg.kg- i given one minute before tracheostomy may be required (Figure 3). extubation may prevent or attenuate ]aryngospasm. ~6 Several other therapeutic modalities have been proposed However, the effectiveness of lidoeaine in this regard is for the treatment of laryngospasm, including hypoxia, controversial. 25'26 Leicht et al. z6 suggest that intrave- doxapram, diazepam and chest compression. +'21-24 We nous lidocaine does not prevent laryngospasm if the do not condone these treatments. patients are extubatecl when they begin to swallow. The Roy and Lerman: LARYNGOSPASM IN PAEDIATRICS 97 authors also suggest that to derive any benefit from the Conclusion lidocaine one must extubate before signs of swallowing Laryngospasm is seen most frequently in paediatric activity. The mechanism of action of intravenous lido- anaesthesia. Although laryngospasm has been regarded as caine may be a central interruption of the reflex pathway, a self-limiting complication, statistically unlikely to or a direct peripheral action on the sensory or motor nerve produce severe complications, 5.0/1,000 patients who terminals. Although the effectiveness of lidocaine is still develop laryngospasm have a cardiac arrest. L In less questioned, we believe that in patients who are at high risk severe instances, laryngospasm may be complicated by of developing laryngospasm, lidocaine should be ad- bronchospasm, hypoxia, gastric aspiration, arrhythmias, ministered intravenously slowly over a 30-second period.2~ and delayed recovery. Pulmonary oedema following Lee and Dowries28 proposed the following guidelines laryngospasm has recently been reported as a significant to diminish the likelihood of post-extubation laryngo- comphcation. ~ A high. level of preparation and fore- spasm. "The infant or child, before tracheal extubation, thought can lead to a reduction in morbidity and mortality should open his eyes and mouth spontaneously, move all as a result of laryngospasm. extremities vigorously and resume a normal breathing pattern after a cough." At the completion of surgery the children should remain on the operating room table with References all monitors in place until the child is awake and fulfill~ 10lsson GL, Hallen B. Laryngospasm during anaes- the above criteria. Furthermore, during emergence, some thesia. A computer-aided incidence study in 136,929 anaesthetists avoid disturbing, touching, or stimulating patients. Acta Anaesthesiol Seand 1984; 28: 567-75. the patient until he/she awakens. This "no touch" technique 2 Fink, BR. The etiology and treatment of laryngeal spasm. is believed to prevent or reduce premature coughing or Anesthesiology 1956; 17: 569-77. bucking while the endotracheal tube is in place. This also 3 Keating VJ. Anaesthetic Emergencies. In: General Anaes- allows the child to awaketl fully before extubation and thesia, 2nd Ed., Chapter 25, Volume 2, pp. 522-3. thus may farther reduce the likelihood of extubation Evans FI' and Gray TC (eds), London: Butterworths, 1965. laryngospasm. 4 SuzukiM, Sasaki CT. Laryngeal spasm: a neurophysiologic Another preventative measure is to directly apply local redefinition. Ann Otol 1977; 86: 150-7. anaesthetic to the supraglottic mucosa. Cocaine (four per 5 lkari T, Sasakf CT. Glottic closure reflex: control mecha- cet~t solution) has been effective in reducing the incidence nisms. Ann Otolaryngol 1980; 89: 220-4. of laryngospasm when applied topically to the vocal 6 Sasaki CT. Development of laryngeal function: etiologic cords, arytenoids and false cords at the end CO2 laser significance in sudden infant death syndrome. Laryngo- surgery, z9 scope 1979; 89: 1964-82. Vagolytic drugs have been recommended as premedi- 7 Rex MAE. A review of the structural and functional basis cants in order to prevent laryngospasm. The results of a of laryngospasm and a discussion of the nerve pathways survey in 1978 revealed that 21 per cent of anaesthetists involved in the reflex and its clinical significance in man advocated the use of IM atropine as premeditation in and animals. Br J Anaesth 1970; 42: 891-9. order to prevent laryngospasm. J~ Although the mecha- 8 HoritaA, DilleJM. Observations on the action ofthio- nism of action of IM atropine is unclear, presumably it is pental (penlothal | tm file larkngeal reflex. Anesthesiology achieved through the elimination of undesirable secre- 1955; 16: 848-53. tions. However, a conflicting report found that fine 9 Heard KM. Benefits and hazards of pentnthal anesthesia. incidence of laryngospasm in those patients receiving Anesthesiology 1944; 5: z,48-64. vagolytics in their premedication was twice that in 10 Rurh HS, TovMI RM, Milligan AD , Charleroy DK. Pcn- patients who did not receive such therapy. J Intramuscular tothal sodium: is its growing popularity justified? JAMA vagolytic premedication is no longer a routine at our 1939; 113: 1864-8. institution. Extubation while the patient is still deeply l I Barron DW, Dundee JW. Clin[cat studies of induction anaesthetized is a method advocated by some. While no agents. XVII: relationship between dosage and side effects data are available, it is our impression that the incidence of intravenous barbiturates. Br J Anaesth 1967; 39: 24-30. of laryngospasm may be reduced. However, we are 12 Burstein CL. Effect of some short acting barbiturates on concerned that this practice will lead 1o the premature the potency of the glo:tis. Proc Soc Exper Biol Med transfer of children to the Recovery Room where they 1937; 37:267 71. may be prone to hypoventilation, aspiration or respiratory 13 Rex MAE. The production of laryngospasm in the cat by obstruction. We currently extubate patienls who are still volatile anaesthetic agents. Br 3 Anaesth 1970; 42: 941-6. deeply anaesthetized only when there is a specific 14 Fisher DM, Robinson S, Brett CM, Pergn G, Grego~ GA. indication to do so. Comparison of enflurane, halothane, and isoflurane for 98 CANADIAN JOURNAL OF ANAESTHESIA

diagnostic and therapeutic procedures in children with malignancies. Anesthesiology 1985; 63: 647-50. 15 Cole WH. Laryngeal spasm and so-called tracheal col- lapse: experimental and clinical studies, Arch Surg 1939; 39: 10-27. 16 Baraka A. Intravenous lidocaine controls e• laryngospasm in children. Anesth Analg 1978; 57: 506-7. 17 Liu LM, DeCook TH, Goudsouzian NG, Ryan JF, Lia PL, Dose response to intramuscular succinylcholine in chil- dren. Anesthesiology 1981; 55: 599-602. 18 Cook DR, Wesrman HR, Rosenfeld L, Hemtershot RJ. Pulmonary edema in infants: possible association with intra- muscular succinylcholine. Anesth Analg 1981; 60: 220-3. 19 Mazze Rl, Dunbar RW. lntralingualsuecinylcholinead- ministration in children: an alternative to intravenous and intramuscular routes? Anesth Analg 1968; 47: 605-15. 20 Smith RM. Anesthetic complications. In: Smith RM. Anesthesia for Infants and Children. St. Louis: CV Mosby Company, 1980; pp 602-3. 21 OwenH. Postextubationlaryngospasmabolishedby doxapram. Anaesthesia 1982; 37:1112-4. 22 ThirutGS. Postextuhationlaryngospasm. Correspondence, Anaesthesia 1983; 38: 393. 23 GoonerameRS, Technique for overcoming laryngospasm. Corresportdence, Anaesth Ir~tenslve Care 1983; 11: 271. 24 Pearc'e H. Laryngeal spasm. Cone~pondence, Anaesth Intensive Care 1983; I 1: 389-90. 25 Geflce K, Andersen LW, Frieset E. Lidocaine given intra- venously as a suppressant of cough and laryngospasm in connection with extubafion after tonsillectomy. Acta Anaesthesiol Seand 1983; 27; 11 I-2. 26 Leicht P, Wisborg T, Chraemmer-Jorgensen B. Does intravenous lidocaine prevent laryngospasm after extuba- tion in children? Anesth Analg 1985; 64:1193-6, 27 Lerman J, Kiskis A. Lidocaine attentuates the intraocular pressure response to rapid intubation in chi]dren. Can Anaesth Soc J 1985; 32: 339-45, 28 Lee KICT, Downes JJ. Pulmonary edema seeortdary to laryngospasm in children. Anesthesiology 1983; 59:347-9. 29 Colman MF, Reynolds R. "How I do it" - head and neck and plastic surgery. A targeted problem and its solution, The use of topical cocaine to prevent laryngospasm after general anesthesia on endoscopy procedures~ Laryngo- scope 1985; 95: 474. 30 Mirakhur RK, Clarke RSJ, Du,dee JW, McDonald JR. Anticholinergie drugs in anaesthesia: a survey of their present position. Anaesthesia 1978; 33: 133-8.