FRCA Notes

Tonsillectomy

The curriculum asks for knowledge of "anaesthetic techniques for common ENT procedures" and "...the surgical procedures used to treat [obstructive sleep apnoea]"

Tonsillectomy featured as a CRQ in March 2021 (89.5% pass rate) and was conisdered "easy" by the examiners.

A CRQ on paediatric OSA and adenotonsillectomy (March 2026) saw examiners quip there was "variable knowledge of paediatric OSA definitions and perioperative respiratory adverse events risk factors".

Post-tonsillectomy bleed, the major surgical complication of this procedure, is covered in a separate page.

Resources

  • Adenotonsillectomy is one of the most common surgeries performed worldwide
  • The vast majority are performed in children (∽16% of all paediatric surgeries in the USA), although it is also performed in adolescents and young adults
  • The tonsils and adenoids form part of Waldeyer's ring; they are largest between 4 and 7yrs of age, after which they regress
  • Children presenting for tonsillectomy carry a higher risk of perioperative respiratory adverse events (PRAEs) than those undergoing most other paediatric surgeries

Obstructive sleep-disordered breathing

  • The main indication is obstructive sleep-disordered breathing (oSDB) due to adenotonsillar hyperplasia
  • Adenotonsillectomy normalises sleep study variables in ∽80% of children
  • Risk factors for residual OSA after surgery include:
    • Age >7yrs
    • Obesity (persistent OSA in 33 - 76%)
    • Severe OSA pre-operatively
    • Asthma
    • Trisomy 21
    • African descent

Recurrent tonsillitis

  • The other main indication is recurrent tonsillitis based on the Paradise criteria
  • Surgery is not indicated with fewer than three episodes in the preceding 12 months
  • The evidence base for this is modest at best, with benefit seen mainly in the first year after surgery

  • Some patients may have elements of both pathological processes, or more of the latter but "over-egg" the symptoms of sleep-disordered breathing in order to qualify for surgery

Other indications

  • Peritonsillar abscess (quinsy)
  • Chronic tonsillitis

Surgical technique

  • Extracapsular tonsillectomy removes the tonsil and its capsule; techniques are either:
    • Cold: sharp/blunt dissection; more intra-operative bleeding
    • Hot: electrocautery, coblation, laser; less intra-operative bleeding but more post-operative pain and secondary bleeding
  • Intracapsular (partial) tonsillectomy, or tonsillotomy, leaves the capsule intact; it is increasingly popular due to less pain and bleeding, at the cost of tonsillar regrowth (2 - 6%)

  • Obstructive sleep-disordered breathing is a spectrum:
    • Primary snoring: no apnoeas, hypopnoeas or desaturations
    • Upper airway resistance syndrome: arousals and sleep fragmentation without desaturation
    • Obstructive sleep apnoea (OSA): recurrent partial or complete obstruction with desaturation ± hypercapnia
  • OSA syndrome implies associated end-organ dysfunction; it is a multi-organ systemic inflammatory condition
  • Prevalence of OSA in children is 1 - 5%, but may exceed 50% in syndromic children

Pathophysiology relevant to the anaesthetist

  • The child with OSA has:
    • An upper airway which is more collapsible at less negative pressures
    • Impaired ventilatory drive and control
    • Blunted arousal and airway self-rescue reflexes
  • Recurrent nocturnal hypoxaemia upregulates brainstem μ-opioid receptors, causing increased sensitivity to the respiratory depressant effects of opioids
  • Paradoxically, systemic inflammation and sleep fragmentation may also increase the pain experienced after surgery

Diagnosis

  • Primarily clinical: habitual loud snoring (≥3 nights/week), witnessed apnoeas, restless sleep with neck extension, daytime hyperactivity or poor concentration
  • Polysomnography is the gold standard, but few children have one before surgery due to cost and access
  • The apnoea-hypopnoea index (AHI) quantifies severity; paediatric thresholds are lower than in adults:
Severity Obstructive AHI (events/hr)
Normal <1
Mild 1 - 5
Moderate 5 - 10
Severe >10

  • Overnight oximetry is a cheaper alternative, scored using the McGill oximetry score (MOS)
    • Clusters desaturations according to degree of hypoxaemia
    • A MOS score of 4 is the most severe, and reflects >3 desaturations to an SpO2 of <80%
    • The MOS score is both diagnostic and risk-stratifies for PRAEs

  • Where no sleep study is available, the STBUR questionnaire can be used:
    • Snoring more than half the time
    • Snoring loudly
    • Trouble breathing or struggling to breathe
    • Stops Breathing during the night
    • Wakes Un-Refreshed
  • Three positive answers triples the risk of PRAEs, and all five increases it ten-fold

Respiratory

  • Increased PRAEs: airway obstruction, laryngospasm, desaturation, negative-pressure pulmonary oedema
  • 3.6-fold increased risk of severe asthma

Cardiovascular

  • Pulmonary hypertension, RV hypertrophy and dysfunction; cor pulmonale if long-standing
  • Systemic (particularly nocturnal) hypertension and LV hypertrophy
  • Endothelial dysfunction and early atheroma

Neurological

  • Hyperactivity, inattention and behavioural problems
  • Impaired memory and executive function; poor school performance
  • Daytime somnolence is less common than in adults

Nutritional and metabolic

  • Failure to thrive in younger children
  • Obesity, insulin resistance and dyslipidaemia (metabolic syndrome) in older children

Haematological

  • Polycythaemia in long-standing OSA
  • Increased risk of stroke in children with OSA and sickle cell disease

Perioperative management of the patient undergoing tonsillectomy


History and examination

  • Screen for OSA: frequency and loudness of snoring, witnessed apnoeas, parental concern about breathing during sleep, daytime behaviour
  • Recent (<2 weeks) URTI or LRTI, frequent wheeze or asthma, and passive smoking exposure all increase PRAE risk
  • History of bleeding with previous surgery (replaces routine coagulation studies)
  • Comorbidities: craniofacial syndromes, trisomy 21, neuromuscular disease, obesity, failure to thrive, prematurity, cardiac disease, sickle cell disease
  • Airway examination: tonsillar size, craniofacial features, previous anaesthetic history

Investigations

  • Rarely needed in healthy children; routine polysomnography is not justified
  • Review any sleep studies and plan post-operative care accordingly
  • Consider ECG and echocardiography in high-risk children with severe OSA and comorbidity e.g. trisomy 21, neuromuscular disease, significant obesity
  • Obese adolescents may require investigation for:
    • Metabolic syndrome, including insulin resistance, LFTs and lipids
    • RV dysfunction using TTE if signs of HTN, severe desaturation (<70%) on PSG or right heart strain

Risk assessment

  • Risk factors for PRAEs after adenotonsillectomy include:
    • Age <3yrs
    • Severe OSA: AHI ≥10, SpO2 nadir <80% or peak CO2 >60mmHg on sleep study
    • Failure to thrive
    • Comorbidity: craniofacial, cardiac, neuromuscular, airway anomalies, ASA >2
    • Recent respiratory tract infection or poorly controlled asthma
  • Only half of the children who suffer adverse events have an easily identifiable age or comorbidity criterion
  • High-risk children should be anaesthetised by the most experienced paediatric anaesthetists available

Day-case or admission?

  • Tonsillectomy is commonly performed as a day-case, but overnight admission is recommended for:
    • Age <3yrs
    • Severe OSA
    • Significant comorbidity
    • Social factors e.g. living >1hr from hospital or no private transport
  • Most primary bleeding occurs within 4 - 6hrs, so day-case patients require an extended observation period; this generally limits day-case surgery to morning lists

Premedication

  • Oral analgesic premedication e.g. paracetamol + NSAID ensures efficacy by the end of surgery
  • Benzodiazepine premedication is associated with PRAEs; alternatives include:
    • Half-dose midazolam (e.g. 0.25mg/kg)
    • ɑ2 agonists e.g. intranasal dexmedetomidine or oral clonidine
    • Parental presence at induction
  • Sedative premedication is best avoided in severe OSA

Induction

  • Children presenting for tonsillectomy may have features which make airway management more challenging:
    • Concurrent medical syndromes
    • Obesity
    • Adenotonsillar hypertrophy and pharyngeal collapsibility
    • Frequent URTIs/asthma
  • Even bag-mask ventilation can be difficult; airway obstruction commonly occurs on induction in OSA
  • Inhalational induction is generally preferred in severe OSA
  • Manoeuvres to manage obstruction:
    • Semi-sitting or reverse Trendelenburg position, using tracheal traction to stiffen the airway
    • Nasopharyngeal airway
    • CPAP

Airway device

  • The airway is shared with the surgeon and access is limited once the Boyle-Davis gag is in place
  • Reinforced LMAs are generally preferred in paediatric patients, as they avoid the need for tracheal intubation and NMBAs
    • However, a south-facing ETT inserted under deep volatile anaesthesia + 1mg/kg IV lidocaine can improve airway safety and obviate need for NMBA
  • Adults undergoing tonsillectomy are usually intubated e.g. south-facing RAE tube

Reinforced LMA

Advantages Disadvantages
No soiling of the lower airway with blood Less secure airway
Avoids NMBA May impair surgical access, especially in younger children
Smoother emergence Risk of gastric distension with IPPV
Airway protection until awake
Less airway trauma

Tracheal tube

Advantages Disadvantages
More secure, definitive airway Requires NMBA or deep anaesthesia
Good surgical access Endobronchial intubation or extubation with head movement
Facilitates controlled ventilation Variable protection from soiling (uncuffed tubes)
Awake vs deep extubation dilemma

  • With either device, re-confirm airway patency after the mouth gag is opened

Anaesthetic agents

  • The impaired neuromotor airway function seen in OSA should prompt choice of agents with the least impact on airway collapsibility:
Agent Propensity for airway collapse
Sevoflurane & desflurane +++
Propofol ++
Opioids ++
Midazolam +
Topical lidocaine +
Ketamine +/-
Dexmedetomidine +/-

  • NMBAs and their reversal agents increase upper airway collapsibility and respiratory complications; use with caution
  • Supplemental oxygen reduces obstruction but can mask it
  • Propofol blunts reflex bronchoconstriction and sevoflurane promotes bronchodilation in the reactive airway

PONV

  • Patients are at a higher risk of PONV and associated dehydration
  • Methods to reduce this include:
    • Surgical technique that reduces swallowed blood
    • Generous crystalloid hydration (10 - 30ml/kg)
    • Prophylactic ondansetron (0.15mg/kg, max 4mg) and dexamethasone (0.15mg/kg, max 8 - 10mg)
    • Limiting opioids
    • Avoiding N2O and minimising fasting times

Emergence

  • Aim for smooth emergence with minimal coughing, laryngospasm and bleeding
  • Suction the pharynx under direct vision and remove any throat pack
  • Recover in the lateral position

Monitoring

  • Transient desaturation in recovery is the most common respiratory event
  • 2 - 6hrs of observation may be required to identify a sentinel respiratory event or primary bleeding
  • A single respiratory event warrants a further 60 mins of observation; repeated events require admission
  • Vulnerability to critical PRAEs is highest within the first 12 - 24hrs; increasing sedation commonly precedes respiratory depression
  • Admitted children with OSA require continuous pulse oximetry and sedation scoring
  • ICU/HDU may be required for severe OSA (e.g. AHI >20 or SpO2 nadir <80%), significant neuromuscular disease, syndromic children, difficult airways or prolonged PRAEs

Pain at home

  • Adenotonsillectomy causes moderate - severe pain, peaking over the first 72hrs and persisting beyond 1 week in half of children
  • Dynamic pain on swallowing limits oral intake, causing dehydration
  • A 2026 BJA cohort study described three pain trajectories in children after adenotonsillectomy:
    • Mild and brief (44%)
    • Moderate (37%): resolving by day 7
    • Severe and prolonged (19%): peaking on days 3 - 5 and persisting into the second week
  • Children with higher STBUR scores (more severe oSDB) were more likely to follow the severe trajectory, particularly younger children
  • The trajectory was evident from the pain score on post-operative day 1; children in the severe group had ∽10x the odds of an unscheduled hospital revisit

Discharge

  • Written discharge advice and telephone follow-up are recommended
  • Parents should be educated on regular (not "as needed") simple analgesia, and on recognising excessive sedation if opioids are prescribed, particularly on the first and second nights
  • Analgesia is typically required for 7 - 10 days

  • Inadequate pain control is a common reason for re-presentation to hospital; an analysis of the PAIN OUT infant registry showed 20% of tonsillectomy patients had a desire for more pain relief
  • A multimodal, opioid-sparing approach is recommended; single analgesics provide only weak to moderate benefit
  • The PROSPECT guideline recommends the following:
Timing Recommended
Pre- and intra-operative Paracetamol
NSAID
IV dexamethasone (single dose)
Acupuncture (adjunct)
Post-operative Regular paracetamol
Regular NSAID
Opioid as rescue only
Honey and acupuncture (adjuncts)
If basic regimen contraindicated Pre-operative gabapentinoid
Intra-operative ketamine (children)
Intra-operative dexmedetomidine

  • Second-line agents are limited by side-effects (sedation, dizziness, hallucinations, bradycardia) and may delay recovery after what is a short procedure

Not recommended

  • Local anaesthetic infiltration or topical application: only short-lived benefit, with reports of arrhythmias, sedation, bleeding and severe upper airway obstruction following lateral pharyngeal injection
  • Lidocaine spray, magnesium, tramadol infiltration, oral or topical ketamine: lack of procedure-specific evidence

NSAIDs, steroids and bleeding

  • Several meta-analyses have shown no increase in post-tonsillectomy bleeding with NSAIDs, and most guidelines now recommend them first-line
  • Ketorolac is the exception and should be avoided
  • COX-2 selective agents (e.g. celecoxib, diclofenac) are attractive alternatives with less antiplatelet effect
  • Dexamethasone was once implicated in increased bleeding; its analgesic and anti-emetic benefits are felt to outweigh any risk

Opioids

  • Children with OSA may simultaneously experience more pain and be more sensitive to opioid-induced respiratory depression
  • If used, opioids should be short-acting and titrated, with the dose halved in OSA (e.g. morphine 50mcg/kg IV)
  • Avoid around-the-clock dosing and concomitant sedatives at home
  • Codeine is contraindicated in children undergoing tonsillectomy for OSA due to ultra-rapid CYP2D6 metabolisers developing morphine toxicity; tramadol, oxycodone and hydrocodone are also CYP2D6-dependent
  • Conversely, standard home morphine doses (≤0.05mg/kg) may be inadequate for children in the severe pain trajectory


Immediate Early (<24hrs) Late (>24hrs)
Trauma/burns to eyes, lips, teeth, pharynx, TMJ Pain, including referred otalgia Pain (may be profound)
Compression/dislodgement of airway device by mouth gag PONV (∽21%, range 13 - 32%) Dehydration and poor oral intake
Intra-operative bleeding Primary post-tonsillectomy bleeding Secondary bleeding (peaks days 5 - 7; up to 28 days)
Airway fire Major PRAE (5.8%) e.g. laryngospasm, aspiration, post-obstructive pulmonary oedema Speech disorders (velopharyngeal incompetence)
Emergence delirium Cranial nerve injury: dysphagia, altered taste
Death from PRAE or haemorrhage Persistent OSA; tonsil regrowth
IJV thrombosis; neck infection

Mortality and readmission

  • Mortality is 0.2 - 0.3 per 10,000 in the UK and Canada
  • Two-thirds of deaths are airway-related, typically occurring at home within the first 24hrs; OSA and recent opioid administration feature prominently
  • The remaining third are due to haemorrhage
  • Unplanned return to hospital occurs in 12 - 13% of children; pain is the most common reason, followed by PONV and dehydration in younger children