Paediatric adenotonsillectomy, part 2: considerations for anaesthesia

Learning objectives.

By reading this article, you should be able to:

• •Describe an approach to perioperative risk stratification and discharge planning after adenotonsillectomy (AT).
• •Discuss the unique airway management challenges in children undergoing AT.
• •Explain key aspects of pain management for AT, including core analgesic tenets, modulating factors, and specific controversies.
• •Discuss various methods to reduce postoperative nausea and vomiting after AT.

Key points.

• •Experienced paediatric anaesthetists should care for children at high risk undergoing adenotonsillectomy (AT).
• •The McGill oximetry score can be used for diagnostic purposes and perioperative risk stratification.
• •Discharge planning after AT varies by institution, but should include monitoring of sedation.
• •There is no single best multimodal anaesthetic or analgesic approach for pain after AT.
• •Multiple factors, including pharmacogenomics, influence pain experienced after AT.

This is the second of a two-part series reviewing the perioperative care of a child undergoing elective adenotonsillectomy (AT) to treat either obstructive sleep-disordered breathing (oSDB)/obstructive sleep apnoea (OSA) and/or recurrent tonsillitis. In Part 1, we reviewed surgical perspectives.¹ In this article, we review anaesthetic aspects, including patient optimisation, risk stratification and discharge planning, airway management and an approach to pain management with a discussion of associated controversies.^2,3

Perioperative risk stratification and planning for postoperative discharge

Children scheduled for tonsillectomy at an increased risk of perioperative respiratory adverse events (PRAEs) need to be observed overnight in hospital. The American Academy of Otolaryngology–Head and Neck Surgery (AAOHNS) foundation guideline provides OSA severity, age and comorbidity-based criteria for overnight hospital admission.² It recommends those with ‘severe OSA’ (apnoea–hypopnoea index [AHI] ≥10 h⁻¹ or Spo₂ nadir <80% on polysomnography [PSG]), children aged <3 yrs and those with certain comorbidities should be admitted. However, perioperative risk stratification is not straightforward. Medical history and physical examination alone cannot identify children at high risk of OSA, and questionnaires that predict OSA or PRAEs have not been validated. To screen for OSA, the authors focus on nocturnal symptoms related to snoring frequency (≥3 nights per week?) and loudness (heard through closed door?), duration and frequency of witnessed apnoeas (>3 nights per week?) and parental concern (‘Are you ever concerned about your child's breathing during sleep?’).⁴ If positive, further questions inquire about negative behaviour, lethargy or attention deficit hyperactivity disorder-like impulsiveness and school performance. Beyond failure to thrive, African descent/family history, and exposure to passive smoke, other conditions associated with OSA are detailed in Table 1 in Part 1.¹ As the number of risk factors increases so does the suspicion of OSA. However, the thresholds of positive responses to predict OSA and determine its severity are currently unknown.

It is unclear how age and comorbidities modulate the severity of OSA, and the frequency and severity of PRAEs in patients presenting for AT. Younger children and adolescents have a higher prevalence for more severe OSA, the latter associated with the obesity epidemic. Age ≤3 yrs and an ASA physical status (PS) >2 are independent risk factors for critical PRAEs.^5,6 Finally, recent (<2 weeks) upper or lower respiratory tract infection (URTI or LRTI); frequent wheezing episodes (more than three times annually) or with exercise; asthma and passive smoking exposure should be assessed in determining perioperative risk, in addition to the comorbidities listed in Table 1 in Part 1 given that only half of children who suffer from severe adverse events have an easily identifiable age or comorbidity criterion.^1,7

Risk stratification using PSG or a more convenient means of objective OSA severity assessment (e.g. a validated questionnaire) is likely to be necessary to predict PRAE. Unfortunately, few children scheduled for AT have a PSG or overnight pulse-oximetry-based OSA diagnosis. When available, there is no consensus on the threshold of AHI that defines ‘severe’ OSA or to identify children at risk for a critical PRAE. The positive predictive values of AHI, respiratory disturbance index and measures of abnormal gas exchange for clinically significant PRAEs remain elusive. Recent evidence suggests that oxygen desaturation nadir and peak CO₂ may be more important than AHI in predicting PRAEs in children undergoing AT.⁸ Consequently, overnight pulse oximetry is an increasingly attractive alternative to PSG because it is more convenient, and it can both diagnose OSA (including severity) and provide risk stratification for PRAEs. Although the absence of desaturations does not exclude OSA, the risk for a PRAE is low.

Finally, the provider's level of experience with paediatric anaesthesia, opioid administration and the postoperative monitoring environment can contribute to the risk of PRAEs.^5,6,9 High-risk children should be managed by the most experienced anaesthetists and observed in a monitored setting. Pain management for all children undergoing AT for oSDB should be multimodal and opioid sparing, and should limit exposure to medications that enhance upper airway muscle relaxation (i.e. ‘pharyngeal sparing’). Figure 1 provides a generic example of an algorithm for perioperative risk assessment and discharge planning.¹⁰ Parental history and, if available, audiovisual recordings, objective adenoid assessment and results of diagnostic tests are necessary to attempt to distinguish between primary snoring and OSA; the latter can increase the risk of PRAEs. The age of the child must also be considered. The US AAOHNS tonsillectomy guideline and a recent UK related report recommend overnight admission for age <3 yrs.^2,11 This guideline precludes the treatment of these children in stand-alone ambulatory surgical clinics in the USA, which cannot accommodate overnight stays. However, it should be noted that age <3 yrs does not imply that care must be provided in a paediatric tertiary care setting, particularly when no high-risk factors or extremes in BMI are present. Particularly so, as a number of secondary care community hospitals in the USA, Canada and the UK exist where anaesthetists with specific paediatric training are present and where a step-down or general care area may be available to accommodate children aged between 2 and 3 yrs requiring more intensive monitoring or respiratory support.¹¹ Nonetheless, age ≤3 yrs should be considered a ‘starting point’ for discussion regarding postoperative discharge planning, including admission after AT given that age ≤3 yrs is associated with severe critical events in data from Europe.⁵ In the USA there is inherent heterogeneity in provider experience and practice setting illustrated by extreme variability in the quality of care provided.¹² Thus, it is paramount that decisions concerning admission and postoperative discharge, including monitoring, are at the discretion of local teams with the assurance that the necessary paediatric expertise, equipment, and overnight care facilities are in place.

Fig 1.

Example of risk stratification and discharge planning for elective tonsillectomy in children. ∗Comorbidities include obesity; failure to thrive; craniofacial and genetic syndromes; neuromuscular, respiratory and haematological disorders; congenital heart disease; and diabetes mellitus.

Preoperative investigations

Beyond a full history, physical examination, airway examination and review of comorbidities, preoperative investigations are unnecessary before AT in healthy children. A history of bleeding from previous surgeries can replace coagulation studies. Routine use of preoperative PSG is not justified. However, in the obese adolescent, a PSG and blood tests investigating for insulin resistance, abnormal liver structure and function and evidence of dyslipidaemia are recommended to exclude metabolic syndrome. Preoperative echocardiography is only recommended in those children with signs of right ventricular dysfunction, systemic hypertension or severe desaturation (<70%) during PSG; full cardiac evaluations are only warranted in children at risk for secondary cardiopulmonary complications (e.g. trisomy-21).¹⁰

Measures to reduce postoperative nausea and vomiting

Children undergoing AT are at increased risk for postoperative nausea and vomiting (PONV) and associated dehydration. Postoperative nausea and vomiting is a common reason for hospital readmission, particularly in younger children. The risk factors are a previous history of PONV, motion sickness, female sex, duration of surgery >30 min, use of volatile anaesthetic agents or longer-acting opioids and anticholinesterase drugs.^13,14 It is recommended that a surgical technique that decreases swallowed blood, generous hydration with crystalloids (10–30 ml kg⁻¹), prophylactic i.v. use of a 5-HT₃ antagonist and dexamethasone are used, and opioids should be limited.¹⁵ Ondansetron (150 μg kg⁻¹; maximum 4 mg) and dexamethasone (150–300 μg kg⁻¹; maximum 8–10 mg) are effective in preventing early (≤6 h) and late (>6 h) PONV, respectively.

Airway management

Management of the airway in children undergoing AT to treat oSDB/OSA or recurrent tonsillitis presents unique challenges.^5,6 Otolaryngological surgery has a relative risk (RR) of 1.46 (95% confidence interval [CI]: 1.12–1.89; p=0.005) for a severe PRAE compared with other common paediatric surgeries.⁵Children presenting for AT include those with medical syndromes, frequent URTIs, asthma and pharyngeal collapsibility. This can make even simple bag-mask ventilation under general anaesthesia difficult for the inexperienced clinician.^5,6,16,17

Difficult airway

Children undergoing anaesthesia for AT can have challenging airways because of OSA airway endotypes (see Table 1 in Part 1).¹ Structurally, children with OSA have a smaller upper airway volume. Difficult airway management, particularly glottic visualisation and tracheal intubation, in addition to challenging bag-mask ventilation, is common in children with craniofacial and soft tissue airway endotypes. Overweight and morbidly obese children have difficult perioperative airway management, increased risk of PRAEs and unexpected admissions, which are complicated by drug dosing outside the recommended range. By managing the airway in a semi-sitting or reverse Trendelenburg position, the anaesthetist can utilise the advantage of tracheal tethering to ‘stiffen’ the airway. Careful placement of a nasopharyngeal airway is effective to manage acute airway obstruction in a patient with OSA, even after AT.

Anaesthetic agents, oxygen and airway neuromotor dysfunction

Upper airway neuromotor dysfunction in children with OSA should be considered in the perioperative anaesthetic plan. As discussed in Part 1, the upper airway of the child with OSA is more collapsible at less negative pressures (see Fig. 1 in Part 1), and there is impaired ventilatory drive, airway self-rescue reflexes, and ventilatory control (see Table 2 in Part 1).¹ For these reasons, the choice of medication in children with OSA should attempt to minimise the impact on airway collapsibility (Table 1) and neuromotor dysfunction.¹⁸ Whilst most anaesthetic agents, including opioids, impair ventilatory drive and airway self-rescue reflexes, α₂-agonists and ketamine have the least impact on blunting airway neuromotor function. Premedication with a benzodiazepine has been associated with PRAEs.^5,6,16 Alternatives to the traditional oral midazolam 0.5 mg kg⁻¹ dose include halving the dose, replacing it with an α₂-agonist or providing parental presence at induction of anaesthesia. Neuromuscular blocking agents and their reversal agents are associated with increased respiratory complications, possibly through enhanced upper airway collapsibility and/or reduced hypoxaemic ventilatory drive; these drugs should be used with caution.^5,6,19, 20, 21, 22 Oxygen administration reduces the propensity for upper airway obstruction by decreasing loop gain. Airway assessment on room air whilst asleep is recommended to allow for the easy identification of residual obstruction and/or impending opioid-related respiratory arrest.

Table 1.

Propensity for anaesthetic agents and opioids to promote upper airway collapsibility. ∗Decrease ventilatory drive and blunt airway self-rescue. NMDA, N-methyl-D-aspartate; CNS, central nervous system; GABA, gamma-aminobutyric acid; NMDA, N-methyl-D-aspartate; EMG-electromyogram.

Generic drug name	Airway collapse	Mechanism of action
Sevoflurane∗	+++	CNS GABAA agonist, proportional to dose
Desflurane∗	+++	CNS GABAA agonist
Propofol∗	++	CNS GABAA/NMDA agonist, proportional to dose
Opioids∗	++	↓Pharyngeal neuromotor drive, ?mechanoreceptor
Midazolam∗	+	CNS GABAA agonist, ?proportional to dose
Topical lidocaine	+	↓Dilator/tensor muscles of the pharynx/larynx.
Dexmedetomidine	+/–	CNS α2-adrenergic agonist
Ketamine	+/–	NMDA receptor antagonist; ↑EMG genioglossus (rats)

Influence of systemic inflammation on airway neuromotor function

Repeated nocturnal hypoxaemia, non-restorative sleep and systemic inflammation are central to the airway management challenges in children with OSA. Sleep deprivation or fragmentation appears to enhance pain perception through inflammatory mediators or central pain-modulatory pathways, which may explain higher reported pain in some patients with OSA.²³ In addition, inflammatory cytokines prolong general anaesthetic gamma-aminobutyric acid A (GABA_A) receptor-mediated amnesia after surgery, and similarly may alter the level of consciousness and airway muscle tone with anaesthetic agents and benzodiazepines.^6,16,24 Obstructive sleep apnoea-related systemic inflammation may predispose patients to critical PRAEs. Consequently, there is a renewed interest in the perioperative administration of steroids and NSAIDs.²⁵

Asthma, recurrent upper airway infections and airway reactivity

Upper and LRTIs and asthma are common in children scheduled for AT. Respiratory problems are associated with a three-fold increase in PRAEs compared with healthy children (RR: 3.15; 95% CI: 2.87–3.46).⁶ This risk is higher still in infants (≤3 yrs), those born prematurely (<37 weeks), or who are exposed to passive smoking.⁵ Respiratory problems are a two-fold stronger predictor for critical PRAEs compared with ASA PS. Bronchospasm and laryngospasm occur most commonly in the operating theatre and recovery area, respectively. Propofol and sevoflurane are the best agents to blunt reflex bronchoconstriction and promote bronchodilation, respectively.¹⁶ Fentanyl, α₂-agonists, and sevoflurane are effective and equivalent in effect to blunt reflex bronchoconstriction. Measures to optimise anaesthetic management to mitigate risk of PRAE are outlined in two excellent reviews.^16,17 Highlights include the value of preoperative inhaled salbutamol, avoiding desflurane, and airway management using a laryngeal mask airway (LMA). In children with a history of asthma, the authors routinely use a reinforced LMA rather than a tracheal tube, with minimal impact on surgical exposure. However, the Boyle–Davis mouth gag may compress or displace the device. In the presence of an associated URTI, LMA removal under deep anaesthesia is preferred to reduce the incidence of laryngospasm.

Postoperative monitoring

Guidelines defer to local practice regarding duration, location, and type of monitoring required for children after AT.^2,26 Transient desaturation in the PACU after AT is the most common respiratory event. Although typically 2–3 h, a range of 2–6 h in the PACU may be required to identify a sentinel respiratory event predisposing to subsequent critical PRAE or primary bleeding after AT. Unfortunately, the type and severity of a sentinel respiratory event to predict a critical PRAE are unknown. However, in adult studies, patterns of hypoventilation, apnoea and hypoxaemia are predictive of critical PRAEs after discharge from PACU.²⁷ One specific respiratory event requires a subsequent 60 min of observation, whilst repeated events require admission. Naloxone administration in PACU, higher-than-usual opioid dosing and interaction with other sedating drugs are also associated with critical PRAEs, which commonly occur within 12–24 h after surgery.^7,28 The authors' criteria for admission to the ICU include children who have undergone urgent/emergent surgery, have high oxygen requirements, are ‘syndromic’, have significant neuromuscular disease or severe OSA (AHI >20 or Spo₂ <80%), a prolonged major PRAE and a difficult airway history. The absence of a PRAE in PACU is likely to indicate that elective admission to the ICU is unnecessary. Guidelines recommend ‘continuous cardiopulmonary monitoring’ that include bedside oxygen saturation. Respiratory depression is commonly preceded by increasing sedation.²⁸ Measures to mitigate against a preventable major PRAE include:

• (i)Improve education to monitor for and identify excessive sedation and respiratory depression.
• (ii)Implement institutional policies related to opioid administration and explicitly identify patients with severe sleep-disordered breathing (SDB)/OSA.
• (iii)Incorporate continuous electronic cardiorespiratory monitoring that is preferentially centrally located.²⁸

Pain management

Adenotonsillectomy is associated with moderate-to-severe pain and has significant impact on the children and their families. Pain with swallowing (dynamic pain) is of concern because it limits the ability to eat and drink or ingest analgesics. For tonsillectomy, pain peaks during the first 3 days and persists in 50% of children after 1 week. Tonsillectomy is associated with the greatest amount of pain, followed by tonsillotomy and then adenoidectomy.²⁹ Poor pain control is a common reason for return to a hospital after AT or a visit from a primary care physician. Acute pain management guidelines advocate using an opioid-sparing approach comprising a combination of NSAIDs and paracetamol (acetaminophen) dosed at regular intervals.^2,26,29 If used, opioids should be titrated individually. In North America, codeine is now contraindicated for pain in children after AT owing to concerns of underlying pharmacogenetic variability predisposing to morphine overdose.³⁰ This, combined with opioid-related adverse events, has resulted in limits on the administration of opioids after AT in children. Non-drug methods (e.g. cold food or distraction) to treat pain are not supported by evidence. Key factors modulating the pain experience are reviewed next. The ideal multimodal analgesic approach is unknown.

Pain experience

Many factors affect the pain experienced after AT.³¹ Increased postoperative pain is reported with AT for recurrent tonsillitis compared with SDB, tonsillectomy compared with tonsillotomy, and when a ‘warm’ vs ‘cold’ surgical technique is used. Patients with oSDB/OSA may paradoxically have both a greater pain experience and increased sensitivity to the respiratory depressant effects of opioids.²³ Sleep deprivation, systemic inflammation and preoperative anxiety may explain the former. To reduce anxiety, parental presence at induction is recommended over sedative premedications, which (particularly benzodiazepines) are associated with increased PRAEs.¹⁶ When sedation is required, clonidine or intranasal dexmedetomidine may reduce opioid consumption, but timing of use may be challenging and recovery may be prolonged. Adequacy of discharge instructions, post-discharge support, and access to medications are health-system-related factors affecting pain control.³¹ Written discharge advice and follow-up by telephone are recommended, particularly in regard to OSA and the risk for over-sedation with opioids.³¹ Concerns over issues of opioid safety, overdose, addiction and the risk of diversion (e.g. oxycodone) should be considered in the care plan. Our centre has established the use of a morphine elixir (1 mg ml⁻¹) with a volume limited to 20 doses (0.1–0.3 mg kg⁻¹ dose⁻¹; maximum 10 mg) as a compromise between effective and safe opioid dosing, accessibility and risk for diversion. A locally prepared suspension of celecoxib (10 and 20 mg ml⁻¹) is also available on-site and in regional pharmacies.

Pain medications

In children undergoing AT, the following perioperative analgesic tenets should be considered: drug classes, dose, timing, route and duration of administration, and drug interactions. Analgesics should be given in an opioid-sparing stepwise approach beginning with an NSAID; then paracetamol; followed by an α₂-agonist (optional); and, finally, a low-dose opioid, titrated as needed.²⁹ Opioid dose reduction by half is recommended as the severity of hypoxaemia is associated with respiratory complications in children with OSA. Preoperative oral loading of celecoxib (6 mg kg⁻¹) and acetaminophen (30 mg kg⁻¹) is recommended. NSAIDs, particularly cyclooxygenase-2 (COX-2) selective drugs, are most effective as a pre-emptive analgesic. Preoperative intranasal fentanyl (1–2 μg kg⁻¹; 50 μg ml⁻¹) or dexmedetomidine (1–2 μg kg⁻¹; 100 μg ml⁻¹) can be used for both sedative and analgesic effects. Intraoperative analgesics can be administered rectally (e.g. paracetamol and diclofenac), intravenously, topically or locally infiltrated into the peritonsillar bed. The most commonly used i.v. opioids are fentanyl 1–2 μg kg⁻¹, morphine 50–100 μg kg⁻¹, oxycodone 50–100 μg kg⁻¹ and tramadol 1 mg kg⁻¹. A single i.v. dose of dexamethasone 0.25–0.3 mg kg⁻¹ (maximum 8–10 mg) can safely improve postoperative pain. Intraoperative ketamine (0.1–0.5 mg kg⁻¹ i.v.) and dexmedetomidine (0.3–0.5 μg kg⁻¹) have complementary cardiovascular effects and attractive opioid and pharyngeal muscle-sparing qualities in patients with severe SDB/OSA, although the effects are short-lived. Dexmedetomidine 1 μg kg⁻¹ is comparable with morphine 100 μg kg⁻¹ for analgesic potency.

In the recovery room, we limit our dose of morphine i.v. to 25–50 μg kg⁻¹. After surgery, proactive celecoxib 3–6 mg kg⁻¹ every 10–12 h (maximum 6–12 mg kg⁻¹ day⁻¹) and paracetamol 15–20 mg kg⁻¹ every 4–6 h (maximum 75–90 mg kg⁻¹ day⁻¹ or 3 g day⁻¹) are prescribed. NSAIDs are associated with less PONV and are more effective in treating dynamic swallowing pain compared with paracetamol.³² Combined NSAID and paracetamol results in analgesic synergism, but no increased toxicity compared with staggered dosing.³³ However, as explained next, NSAIDs may predispose to an increased risk of postoperative haemorrhage. Although paracetamol-associated hepatotoxicity after AT has not been reported, dose reduction is recommended by the third day. An opioid in the evening may improve overnight sleep, but the child should be assessed for excessive sedation on the first and second nights.³³ Analgesic administration is typically for 7–10 days with tapering of drugs in reverse order of their administration.

Analgesia-related controversies

NSAIDs, steroids and risk of post-tonsillectomy bleeding

The US surgical paediatric tonsillectomy guideline strongly recommends ibuprofen (5–10 mg kg⁻¹ every 6–8 h) to treat pain after AT.² However, NSAIDs predispose to an increased risk for postoperative haemorrhage after AT. Specifically, all NSAIDs inhibit two COX enzymes to varying degrees: COX-1, a body maintenance enzyme that is responsible for integrity of the gastric mucosal barrier and platelet aggregation/vasoconstriction and COX-2, an inducible enzyme that is also responsible for gastric mucosal and renal cytoprotection. The contribution of COX-1 and -2 to prostanoid-related inflammation and pain is unknown. Available NSAIDs inhibit primarily COX-1 (e.g. ketorolac), COX-2 (e.g. celecoxib>diclofenac), or both COX-1 and COX-2 (e.g. ibuprofen, naproxen and indomethacin). Quantifying the risk of bleeding associated with NSAID use is difficult because of the heterogeneity of studies (e.g. surgical indications) and insufficient sample size. Ibuprofen is comparable with celecoxib and diclofenac for risk of gastric erosion, whilst naproxen and indomethacin, and particularly ketorolac, carry higher risks. Ibuprofen is associated with increased bleeding severity, and therefore, a lower dose combined with acetaminophen or delaying administration until after surgery is recommended.^2,33 Ketorolac has been associated with excessive bleeding in clinical trials.² Although an intraoperative dose of dexamethasone can increase the severity and frequency of bleeding, the benefits of dexamethasone outweigh the risk.¹⁴ Given the controversy of NSAIDs and bleeding after AT, celecoxib (or diclofenac) is an attractive alternative analgesic. It has a wide therapeutic index and COX-2-specific selectivity, and is not contraindicated in patients with sulpha allergies. Our published experience of celecoxib and intraoperative dexamethasone is the primary reason for adopting a celecoxib suspension in our patients undergoing AT and at the Children's Hospital of Philadelphia (https://clinicaltrials.gov/: NCT02934191).³⁴ Selective COX-2 inhibitors are safe in patients in whom NSAIDs are contraindicated because of allergy, asthma or NSAID-exacerbated respiratory disease.

Opioids and risk for postoperative respiratory depression in children with OSA

There appear to be two competing issues regarding administering opioids in children with OSA: heightened sensitivity to the analgesic and respiratory effects of opioids, and evidence for increased pain after AT.²³ There is a growing concern of both coexisting in the same patient with OSA, increasing their risk of an opioid-related PRAE.²³ In animal models, repeated hypoxaemia is associated with neuroplastic changes and central brainstem upregulation of the μ-opioid receptor. The molecular processes of this receptor gene link opioid analgesia and opioid-induced respiratory depression. Hypoxaemia in OSA is associated with both reduced opioid requirements and respiratory depression with opioids in children.²³ Both opioids and OSA have been implicated in cases of anoxic brain injury or death after AT in children.⁷ The pharmacogenetic variability of codeine highlights only one aspect of this risk.³⁰ Racial disparity in pain management for children of African descent compared with Caucasian after AT may simply reflect a higher prevalence of OSA or underlying pharmacogenetic differences in the former.³⁵ Acute opioid analgesic tolerance, associated hyperalgesia and opioid dosing in the obese patient complicate strategies to minimise the risk for opioid-related major PRAEs.

There is no ideal opioid to use for a child undergoing AT, as all opioids impair ventilatory drive and airway self-rescue, and increase pharyngeal collapsibility during sleep. The narrow therapeutic index of opioids dictates reducing the initial dose by 50% whilst avoiding around-the-clock dosing and concomitant sedatives. Better education of care providers and parents of the inherent risks of opioid-induced respiratory depression and need for improved postoperative assessment in the first 12–24 h after surgery is recommended.²⁸ Although oxycodone may be more favourable in terms of palatability, potency, analgesic efficacy and gastrointestinal-related adverse effects, most Canadian paediatric tertiary care centres prescribe oral morphine because of its easy access, lower risk of drug diversion and fairly consistent analgesic potency. Doses of 100–300 μg kg⁻¹ (usually ≤200 μg kg⁻¹; maximum 5–10 mg dose⁻¹) are given every 3–4 h as needed in children 2–12 yrs old.

Topical compared with local infiltration of local anaesthetics and other adjuvant drugs

Local anaesthetics, dexamethasone, α₂-agonists, ketamine, opioids and magnesium have been investigated for their analgesic effects when used either topically or infiltrated at the tonsillectomy site.^14,36 Local anaesthetics appear to provide only a modest reduction in resting pain, lasting up to 24 h. Increased pharyngeal collapsibility may be an unintended consequence. Similarly, locally infiltrated dexamethasone appears to reduce early postoperative pain, total analgesic and anti-emetic consumption.¹⁴ Our current practice does not include this technique.

Genetics and pain management

Genetic variations influence the pain experienced and care of the child undergoing AT. A child's pharmacogenomics affects the pharmacokinetics and pharmacodynamics of analgesic drugs, and their growth and development influence genetic expression (i.e. ontogeny). Other factors also influence the perception of pain, possibly via epigenetic mechanisms, including catecholamines, inflammatory cytokines, race, sex, diet, exercise and other drugs. However, it is estimated that genetics accounts for <2 cm of variability in pain perception on a 10 cm pain scale; genetic susceptibility to adverse events may be more important. Some genetic factors of interest and relevant to pain management in children undergoing AT are the pharmacogenomics of opioids and NSAIDs and race. Ultra-rapid metaboliser liver cytochrome P450 family CYP2D6 genotypes have led to concerns about analgesic efficacy and safety, limiting or banning the use of codeine outright in children.³⁰ Opioid alternatives include hydrocodone, oxycodone and tramadol, but they all have CYP2D6 metabolites that have greater opioid potency than the parent drug in contrast with morphine.³⁰ Each of these drugs has been implicated with overdose and associated respiratory depression or inadequate pain relief. Importantly, concomitant use or discontinuation of CYP3A4 inhibitors (e.g. clarithromycin) can increase hydrocodone and oxycodone plasma concentrations leading to fatal respiratory depression. Single-nucleotide polymorphisms (SNPs) for genes associated with morphine pharmacokinetics (e.g. OCT1, UGT2B7, and ABCB1/ABCC3) help explain paediatric variability in efficacy and associated adverse events, including observed racial disparities. The SNPs of the CYP2C9 enzyme may explain improved NSAID analgesia.^3,34 However, tailoring opioid and NSAID analgesia based on selective genotyping with the possible exception of the CYP2D6 genome is not currently justified.³

Conclusions

The perioperative anaesthetic care of a child undergoing AT is evolving. Perioperative risk stratification is becoming less reliant on PSG as overnight pulse oximetry is an increasingly acceptable alternative. Validated paediatric questionnaires to diagnose OSA or predict critical PRAEs are being developed. These are likely to incorporate objective measures, such as overnight pulse oximetry, to improve performance of the tools. Preparation is required to address the unique airway challenges inherent in the child undergoing AT. Increased pain and sensitivity to opioid-induced respiratory depression in the child with OSA compel using a multimodal opioid-sparing analgesic technique. However, there is no one ‘best’ multimodal anaesthetic or analgesic approach to treat pain after AT. Vulnerability to critical PRAEs is highest within the first 12–24 h after AT, warranting better education of attending providers and parents, and implementing measures to identify excessive sedation and respiratory depression. It is apparent that providers with the most experience in paediatric anaesthesia should care for high-risk patients undergoing AT.

Declaration of interests

The authors declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Acknowledgements

The authors would like to thank Drs Debbie Schwengel and Britta S. von Ungern-Sternberg for their feedback regarding perioperative discharge planning and management of the sensitive airway, respectively. The authors would also like to thank Johanna Spaans for editing the manuscript.

Biographies

Julie Zalan MD FRCPC is an anaesthetist with specialty training in paediatrics working at Kingston Health Sciences Centre, and an assistant professor at the Department of Anesthesiology and Perioperative Medicine, Queen's University, Kingston.

Jean-Philippe Vaccani MD FRCPC is a paediatric otolaryngologist working at the Children's Hospital of Eastern Ontario (CHEO), and an associate professor and vice-chair of education at the Department of Otolaryngology–Head and Neck Surgery, University of Ottawa. He has published previously on paediatric obstructive sleep apnoea.

Kimmo Murto MD FRCP is a paediatric anaesthetist working at the CHEO, and an associate professor at the Department of Anesthesiology and Pain Medicine, University of Ottawa. He is past chair of the paediatric committee of the Society of Anesthesia and Sleep Medicine. His research interests include care of children undergoing tonsillectomy.

Matrix codes: 1H02, 2D02, 3A02