Papers [Next][Papers][Home Page][Previous]
Paper 35  

Anaesthesia for Tonsillectomy in Children

Dr Craig Sims FANZCA.
Staff Anaesthetist.

Dr Chris Johnson FANZCA.
Visiting Specialist.

Department of Anaesthesia,
Princess Margaret Hospital for Children,
Western Australia.

From AUSTRALASIAN ANAESTHESIA 1996 with the permission of the Australian & New Zealand College of Anaesthetists and the editor Dr John Keneally.

* History Repeats Itself * Local Anaesthesia for Pain Relief
* Surgery * Postoperative Vomiting
* Anaesthetic Technique * Summary
* Laryngeal Mask Airway for Tonsillectomy * Acknowledgement
* Obstructive Sleep Disorder * References
* NSAIDs and Analgesia  
History Repeats Itself

Approaches to tonsillectomy during the early years at the Children's Hospital in Perth evolved and changed in a manner not unlike that which is occurring now. Tonsillectomy was definitely considered an outpatient procedure, particularly as an operation on the kitchen table at home was one of the options prior to the opening of the hospital in 1909. It is interesting to note that there were no intraoperative deaths recorded during tonsillectomy in the 13 years before 1924 (some 550 patients).

The father of one of the authors remembers travelling home by public transport after outpatient tonsillectomy at age 4 during 1925. Home was in a then rather remote location outside Perth, with no possibility of a return journey in the event of complications or bleeding. It was not until 1938 that overnight admission was considered, partly in an attempt to reduce the infection rate and need for subsequent inpatient readmission, and partly for social reasons. The Hospital Board appreciated that residents of the suburb near the hospital had developed the habit of avoiding late afternoon trams which were often full of small, unwell patients returning home after their tonsillectomy.

The approach to airway management also varied with time. Anaesthesia using an endotracheal tube was first recorded at the Children's Hospital in Perth in 1924 for antral surgery. Tubes were not commonly used for tonsillectomy until 1927. By 1930, tubes were used reasonably frequently, but when the use of intrapharyngeal anaesthesia was first recorded in 1932, followed by the Boyle Davis gag in 1936, the use of endotracheal tubes declined rapidly. By 1956, one particular ENT surgeon would not allow intubation to be performed for tonsillectomy. It was not until the 1960s, that endotracheal anaesthesia became the norm for tonsillectomy, due to the insistence of the new Director of Anaesthesia, Dr Nerida Dilworth.

Tonsillectomy was performed on 2486 children in Western Australia during 12 months in 1994/95. Despite the enormous numbers of tonsillectomies performed each year, we still do not know the best anaesthetic technique to reliably give satisfactory operating conditions, rapid theatre turnover, safe awakening and a happy, comfortable child on the ward and at home. Recent changes have included renewed interest in tonsillectomy as a day stay procedure, avoidance of endotracheal intubation by using the reinforced laryngeal mask, and an increasing number of children having surgery for obstructive sleep disorder. [Top]


Surgery tends to be for recurrent infections in older children, and for chronic airway obstruction in preschool aged children. No laboratory tests are required routinely before tonsillectomy. Less than 0.5% of healthy children have a preoperative haemoglobin less than 100g/l, and less than 0.1% of children have a clinically important coagulopathy. The guillotine operation has essentially been abandoned, and now either a dissection/snare followed by point coagulation, or electrocautery excision, is used. Adrenaline soaked gauze, bismuth paste or laser coagulation are sometimes used to control bleeding.

An average of 5% of the blood volume is lost during surgery; hence IV fluids should be given routinely. Blood loss can be much larger and about 1 in 20 children lose more than 10% of their blood volume. Care must be taken with the type of IV fluid used to replace this degree of blood loss, as hyponatraemia is possible if glucose-saline fluids are used. Blood transfusion is needed in less than 0.1% of children. This low use of blood is not surprising, because if normovolaemia is maintained with asanguinous fluid, nearly half of a child's blood volume can be lost before the haemoglobin reaches 70g/l.

Around 1% of children have significant postoperative bleeding; some of these patients need to return to theatre for further treatment. In this circumstance, anaesthesia requires resuscitation with colloid or Hartmann's solution and a rapid sequence induction with cricoid pressure. The dose of the induction agents should take into account the possibility of unrecognised hypovolaemia.

Occasionally, a coagulopathy is first detected at the time of tonsillectomy. Von Willebrand's disease (reduced von Willebrand factor (vWF), leading to abnormal platelet function) is the most common coagulopathy found in this situation. DDAVP 0.3 micrograms/kg given intravenously over 30-60 minutes, releases stored vWF from capillary endothelial cells and improves haemostasis.

Tonsillectomy on a day stay basis has been limited, because of concerns about postoperative haemorrhage, vomiting and upper airway obstruction. Haemorrhage usually occurs in the first six hours postoperatively, but may occur later, following discharge of a day patient. Continued bleeding may be unrecognised, if the child swallows blood without vomiting. Furthermore, there are concerns regarding transferring the relatively difficult and at times unpleasant postoperative care to parents. Children having adenoidectomy alone have a lower complication rate and any significant bleeding tends to occur in the early postoperative period. For these reasons, adenoidectomy is often done as a day surgery procedure, with at least 6 hours of observation between surgery and discharge. [Top]

Anaesthetic technique

Any anaesthetic technique for tonsillectomy should result in a child who is conscious on arrival in recovery and has good postoperative analgesia, without excessive sedation. A screaming patient in recovery is not necessary for safety, is unsatisfactory for parents and recovery room staff, and may be more likely to continue bleeding from the tonsillar bed. Laryngospasm and stridor after extubation in tonsillectomy patients has been reported to occur in between 0% and 20% of patients 1. The effect of various methods and manoeuvres have been studied, but it seems the care and skill of the anaesthetist is more important than any particular technique.

An anaesthetic technique based on spontaneous ventilation with deep extubation has the advantage of allowing rapid turnover of cases with a quick surgeon and perhaps reduces the risks associated with accidental disconnection. This technique does, however, have several major disadvantages. Most importantly, responsibility for managing upper airway obstruction during emergence is transferred to the recovery room staff. This is especially likely to be hazardous in children with obstructive sleep disorder (now the commonest indication for tonsillectomy in children), because upper airway obstruction during emergence is likely to be more troublesome. In the authors' experience, urgent calls for assistance in recovery are not infrequently related to use of this technique. The risks of obstruction, laryngospasm and desaturation are greatly reduced in a conscious child.

Other criticisms of this technique include the elective use of suxamethonium for intubation and hypercarbia during spontaneous respiration, resulting in ventricular arrhythmias (incidence 20% with halothane). Finally, titrating opioid dosage may be more difficult during spontaneous ventilation.

An anaesthetic technique based on positive pressure ventilation and awake extubation places the patient in the care of the anaesthetist at the time when upper airway obstruction is most likely. Arrhythmias are less frequent using this technique and perhaps it is easier to give an adequate dose of opioid intraoperatively. However, care must be taken with doses of nondepolarising relaxant, particularly with a quick surgeon. Reasonable intubating conditions can be achieved using atracurium 0.25 mg/kg, supplemented by opiate and propofol. Knowledge of the surgeon's operating habits can prevent delays caused by misjudgment of times for awakening and extubation. Sevoflurane and mivacurium may also have a useful role.

Despite the frequency of anaesthesia for tonsillectomy in children, there are few studies comparing different anaesthetic techniques or agents commonly in use. Anaesthetists should consider the techniques and agents available, the increasing number of children with obstructive sleep disorder presenting for surgery and the increasing patient and parent desire for a relatively smooth and pleasant postoperative course. [Top]

Laryngeal Mask Airway for Tonsillectomy

The reinforced laryngeal mask airway (rLMA) has recently presented an alternative method of maintaining the airway without endotracheal intubation 2. It has an armoured, narrow bore tube which can be flexed and taped to the centre of the chin. The tube fits under the split tonsillar gag as for a RAE tube. The gag holds the rLMA in position and the cuff is not visible in the mouth. Anaesthesia for tonsillectomy using the rLMA has been described for both adults and children. Advantages include avoidance of complications related to laryngoscopy and endotracheal intubation. In particular, complications related to intubation in children who have had a recent upper respiratory tract infection may be avoided. There is also a reduction in the incidence of tracheal aspiration of blood from 54% to 10% with the ETT and rLMA respectively, and a quieter emergence from anaesthesia with less airway obstruction.

The problems associated with the use of the rLMA for tonsillectomy include difficulties with insertion or positioning, leading to about a 10% failure rate. Opening the gag may also push the mask anteriorly and cause the aperture to be occluded by the epiglottis. A technique based on spontaneous ventilation will be difficult if occlusion or malposition is present. Maximal advantage of the rLMA would be achieved by using a spontaneous ventilation technique with awake removal in recovery. In this way, a fast turnover in theatre is achieved, and the airway protected from blood until the patient is awake.

The ultimate role of the rLMA for tonsillectomy depends on the experience of the anaesthetist, the availability and cost of the rLMA, and the willingness of the anaesthetist and recovery staff to adopt new techniques. [Top]

Obstructive Sleep Disorder.

Obstructive sleep disorder (OSD) is characterised by upper airway obstruction during sleep, leading to hypoventilation and sleep disturbance 3, 4. It is not a form of adult obstructive sleep apnoea, but a separate syndrome with specific symptoms, diagnostic criteria and treatment. It occurs particularly in children between 2 and 5 years old (when pharyngeal lymphoid hyperplasia is maximal) and, in contrast to adults, there is no male predominance. In children, adenotonsillar hypertrophy is the major risk factor for OSD; in adults, it is obesity. OSD occurs in about 1-2% of 2-5 year olds and is now the main indication for adenotonsillectomy in children.

Airway obstruction during sleep is rare in normal children. However, respiratory pauses which are less than 10 seconds long occur during sleep in children of all ages. In contrast to adults, children with OSD often manifest a pattern of persistent partial airway obstruction, rather than cyclical, episodic obstructive apnoeas. However, the severity of OSD does not always correlate with the size of the tonsils and adenoids and it is not known whether tonsillar or adenoidal size is most important. It is also not known why only some children with adenotonsillar hypertrophy develop OSD. Unlike adults, children with OSD appear to have normal central control of upper airway muscles during sleep. However, inspiratory movement of the enlarged tonsils and pharyngeal walls tends to narrow the upper airway and occasionally cause obstruction. The upper airway is also narrowed in children with craniofacial anomalies associated with midfacial or mandibular hypoplasia, such as Down and Pierre Robin syndromes, predisposing them to OSD.

Snoring is a regular finding in children with OSD. Other symptoms are nocturnal sweating, restless sleeping with many changes of position, periods of apnoea and, occasionally, enuresis. Breathing while awake is almost always normal, but morning headaches, hyperactivity, failure to thrive, mouth breathing and changes in facial growth may occur. Daytime somnolence is much less common than in adults.

The episodes of nocturnal hypoxaemia may rarely result in significant pulmonary hypertension and impaired function of left and right ventricles. Increased awareness of abnormal breathing during sleep has resulted in a decline in the number of children who develop severe pulmonary hypertension.

Adenotonsillectomy is usually an effective treatment for OSD. Even if there is an underlying craniofacial abnormality, adenotonsillectomy is often the first surgical treatment tried. Both airway symptoms and cardiovascular changes improve after surgery. Anaesthesia concerns relate to postoperative airway obstruction and the presence of pulmonary hypertension. Parents of a child who snores at night should be asked about other symptoms of OSD. Such a child who also has apnoeas noticed by the parent should be considered to be in a high risk group and managed differently from the child having tonsillectomy for recurrent infections.

Postoperative airway obstruction due to residual anaesthetic agents, opioid analgesics and surgical oedema are all concerns. However, there is little evidence of superiority of one anaesthetic technique over another. Nevertheless, a technique which reduces postoperative sedation as much as possible is generally used. Methods used include avoidance of heavy premedication, minimal use of volatile anaesthetics, care with opioid dose and a relaxant technique. The principle aim is to ensure smooth, awake extubation without respiratory obstruction. Postoperative care should be in an area with continuous observation, including pulse oximetry. Nursing staff should be skilled in the observation and care of children who have had tonsillectomy, ensuring that problems in the postoperative period will be detected early and managed appropriately until further help arrives. Many centres acknowledge the relatively high risk of surgery and anaesthesia by managing some or all of this group of patients in the paediatric intensive care unit. Children aged less than 3 years, or with craniofacial abnormalities, are a particularly high risk group for postoperative airway obstruction. [Top]

NSAIDS and analgesia.

NSAIDs are frequently used as part of a multi-modal approach to managing postoperative pain, and reduce morphine consumption by about 30% after a variety of procedures. Diclofenac and ketorolac provide analgesia similar to opioids, in tonsillectomy patients. Their potential advantages include a reduction in sedation, respiratory depression and emesis. However, postoperative restlessness is more common and there is no direct evidence that tonsillectomy patients are better off receiving NSAIDs rather than opioids. Furthermore, there is concern that these drugs may increase the risk of perioperative bleeding due to their effects on platelet function 5.

It is not known with certainty whether or not NSAIDs affect blood loss after tonsillectomy. Studies (mostly on small numbers of patients) have suggested both unchanged and increased perioperative blood loss in tonsillectomy patients given NSAIDs during surgery. Changes in perioperative bleeding are difficult to study, because it is not easy to assess intraoperative blood loss and even harder to measure it postoperatively. The incidence of postoperative bleeding requiring return to theatre is so low that a very large study would be needed to prove a difference with and without NSAIDs.

NSAIDs may have a significant role in the subgroup of patients with obstructive sleep apnoea syndrome, where avoidance of excessive sedation is important. At present however, there is inadequate evidence to define their role in analgesia. [Top]

Local Anaesthesia for Pain Relief

Peritonsillar infiltration with local anaesthetic should have the potential to reduce pain after tonsillectomy, but studies have been small and mostly demonstrated analgesia of less than 1 hour 6 . Solutions containing adrenaline may halve blood loss, but the risk of injection into the internal carotid artery has always been a concern. Spraying local anaesthetic onto the tonsillar bed after removal of the tonsils is not effective. [Top]

Postoperative Vomiting.

Vomiting is an important, common and unpleasant occurrence after tonsillectomy. Up to 75% of children vomit after tonsillectomy, and 10% vomit 3 or more times. Propofol anaesthesia, metoclopramide 0.15- 0.25mg/kg, ondansetron 0.15mg/kg (oral or intravenous), tropisetron 0.1mg/kg and droperidol 0.05mg/kg are all effective, to varying degrees, in reducing vomiting after tonsillectomy 7.

Swallowed blood contributes to the high incidence of vomiting after tonsillectomy. Some concerns have been raised over anecdotal reports of antiemetic therapy suppressing vomiting, thereby preventing recognition of continuing postoperative haemorrhage. No information is available on the effectiveness of antiemetics when the stomach is full of blood. Postoperative haemorrhage occurs in less than 1% of cases, and retching and vomiting may even contribute by causing venous engorgement. If antiemetics are deliberately avoided so as to not to conceal postoperative haemorrhage, many children will be left to suffer and vomit after tonsillectomy.

Prophylactic antiemetic therapy should be considered for all children over the age of 2 years undergoing tonsillectomy. [Top]


Attitudes to tonsillectomy have slowly changed over the years. A Board minute from the Perth Children's Hospital in May 1932 reads :

"Permission is given to go to Parkerville Home (an orphanage) on some convenient Sunday. The Medical Superintendent is to take with him a junior doctor and the necessary anaesthetics for the purposes of removing T & As from a large batch of children."

This relaxed attitude is not acceptable nowadays. Major morbidity is now very rare, but an increasing proportion of our patients are at high risk of postoperative upper airway obstruction. Pressure to perform day stay tonsillectomy must be balanced by the need for adequate observation after surgery. Factors causing minor morbidity such as pain and vomiting must also be addressed if the often unpleasant task of post operative care is to be delegated to parents.

Although tonsillectomy is a common procedure, there is little knowledge available based on scientific data to make choices about anaesthetic techniques. Auditing one's own techniques and keeping an open mind about new drugs and methods are vital. [Top]


The authors would like to thank Dr N Dilworth and Ms J Marshall for the historical information used in this article. [Top]


  1. Klowden AJ, Rifai S, Salem MR. Can post-extubation laryngospasm after tonsillectomy and adenoidectomy in children be prevented? Anesth Analg 1995; 80:S239
  2. Williams PJ, Bailey PM. Comparison of the reinforced laryngeal mask airway and tracheal intubation for adenotonsillectomy. Brit J Anaesth 1993; 70: 30-33
  3. Gaultier C. Obstructive sleep apnoea syndrome in infants and children: established facts and unsettled issues. Thorax 1995; 50: 1204-10
  4. Helfaer MA, Wilson MD. Obstructive sleep apnea, control of ventilation, and anesthesia in children. Pediatr Clin N Am 1994; 41:131-151
  5. Thiagarajan J, Bates S, Hitchcock M, Morgan-Hughes J. Blood loss following tonsillectomy in children. A blind comparison of diclofenac and papaveretum. Anaesthesia 1993; 47: 132-5
  6. Wong AK, Bissonnette B, Braude BM et al. Post-tonsillectomy infiltration with bupivacaine reduces immediate postoperative pain in children. Can J Anaesth 1995; 42 :770-4
  7. Rose JB, Martin TM. Post tonsillectomy vomiting. Ondansetron or metoclopramide during paediatric tonsillectomy: are two doses better than one? Paediatr Anaesth 1996; 6:39-44 [Top]

[Guidance Notes] [Keyword Search] [Papers] [Reference] [Home Page]