Anesthesia for tonsillectomy with or without adenoidectomy in children

INTRODUCTION — Tonsillectomy with or without adenoidectomy is one of the most common surgical procedures performed in children. Airway obstruction, resulting in sleep-disordered breathing or obstructive sleep apnea (OSA), and recurrent infection are the most common indications for adenotonsillectomy. Forty years ago, the majority of tonsillectomies were performed for infection, while OSA is now the most common indication, accounting for over 75 percent of cases.

The following features make airway management for adenotonsillectomy particularly challenging:

●The airway is shared between the anesthesiologist and the surgeon and must be protected from blood and secretions.

●The incidence of laryngospasm is higher than during other surgical procedures.

●Children who undergo the procedure for OSA are at particularly high risk of significant respiratory complications in the postoperative period.

This topic will discuss the anesthetic management of children who undergo tonsillectomy with or without adenoidectomy, as well as the management of the child with a hemorrhage after tonsillectomy. Evaluation of sleep-disordered breathing and OSA in children, as well as indications for tonsillectomy, are discussed separately. (See "Evaluation of suspected obstructive sleep apnea in children" and "Management of obstructive sleep apnea in children", section on 'Surgical therapy' and "Tonsillectomy and/or adenoidectomy in children: Indications and contraindications".)

PREANESTHETIC EVALUATION — Preoperative evaluation in anticipation of adenotonsillectomy should pay particular attention to the likelihood and severity of obstructive sleep apnea (OSA) and of bleeding disorders. Upper respiratory infections (URIs) are common in children presenting for adenotonsillectomy and may affect the timing of surgery, as recent URI increases the risk of perioperative adverse respiratory events [1]. This is discussed in more detail separately. (See "Anesthesia for the child with a recent upper respiratory infection" and "Tonsillectomy and/or adenoidectomy in children: Indications and contraindications", section on 'Infectious'.)

Screening for obstructive sleep-disordered breathing and OSA in children and assessment of severity of OSA are discussed separately. (See "Adenotonsillectomy for obstructive sleep apnea in children", section on 'Initial screening' and "Evaluation of suspected obstructive sleep apnea in children", section on 'Screening' and "Evaluation of suspected obstructive sleep apnea in children", section on 'Assessment of severity'.)

Bleeding disorders in children are also discussed separately. (See "Approach to the child with bleeding symptoms".)

ANESTHETIC MANAGEMENT FOR ADENOTONSILLECTOMY — A variety of anesthetic techniques are used for tonsillectomy. Overarching goals of anesthetic management include:

●Smooth, atraumatic induction of anesthesia

●Protection of the airway during surgery

●Provision of postoperative analgesia

●Prevention of postoperative nausea and vomiting (PONV)

●Smooth, rapid emergence to allow recovery of airway protective reflexes and avoidance of airway obstruction and respiratory depression

Anesthetic strategy — The strategy for anesthesia for tonsillectomy should be developed by the anesthesiologist and the surgeon, considering institutional staffing, resources, and protocols. There is wide variability in anesthetic management for tonsillectomy. The following is our approach:

●Premedication – For most children undergoing tonsillectomy, we use distraction techniques (eg, videos, songs, mobile computing devices) and parental presence during induction of anesthesia rather than premedication with anxiolytics. The goal is to calm the child while avoiding the sedative and respiratory depressant effects of anxiolytics. (See 'Preparation for anesthesia' below.)

●Induction technique – For young children (≤6 years of age) without severe obstructive sleep apnea (OSA), we perform inhalational induction of anesthesia with sevoflurane. For older children (≥12 years of age), we prefer intravenous (IV) induction to allow more rapid induction and airway control. For children between 6 and 12 years of age, the induction technique depends on the child's level of anxiety and ability to tolerate IV placement.

For children of any age with severe OSA, or for children with multiple risk factors for perioperative respiratory adverse events (table 1), we prefer an IV induction whenever possible. If inhalational induction is necessary for these children, IV access should be obtained as soon after induction as possible.

Propofol is the preferred IV agent for hemodynamically stable patients. We routinely apply topical local anesthetic prior to IV placement. (See 'Induction of anesthesia' below and "Clinical use of topical anesthetics in children", section on 'Lidocaine-prilocaine'.)

●Airway management – We prefer a cuffed endotracheal tube (even in very young children) rather than an uncuffed endotracheal tube or laryngeal mask airway for airway management in order to protect the airway from bleeding, secretions, and the risk of fire with the use of electrocautery. (See 'Choice of airway device' below.)

●Neuromuscular blocking agents (NMBAs) – For most children, especially younger children, we routinely intubate without the use of NMBAs. (See 'NMBAs' below.)

●Maintenance of anesthesia – We maintain anesthesia with sevoflurane and a low concentration of inspired oxygen (25 to 30 percent). (See 'Maintenance of anesthesia' below.)

●Analgesia – We use a multimodal approach to pain control, which includes prophylaxis with intraoperative IV acetaminophen (15 mg/kg), as well as an intraoperative dose of dexamethasone (0.1 to 0.5 mg/kg IV, maximum of 4 mg), usually in combination with a low-dose opioid (morphine 0.1 mg/kg and/or fentanyl 1 mcg/kg IV). For many children we administer dexmedetomidine 0.5 mcg/kg IV for its opioid sparing effects and to minimize emergence delirium. Opioid doses are reduced for children with obstructive sleep apnea. (See 'Analgesia' below.)

●Antiemetics – PONV is decreased by the use of intraoperative dexamethasone along with a serotonergic antagonist (eg, ondansetron 0.1 mg/kg IV, maximum of 4 mg). (See 'Antiemetics' below.)

●Emergence – After thorough suctioning of the oropharynx, we extubate in the operating room when the patient is awake and able to protect the airway. (See 'Emergence from anesthesia' below.)

Preparation for anesthesia — Prior to the induction of anesthesia, patient care focuses on the reduction of anxiety, the initiation of analgesics, and proper patient positioning:

Reduction of anxiety — While anxiolytic premedication is often administered prior to anesthesia, we avoid it for children with sleep apnea undergoing adenotonsillectomy whenever possible. Often, distraction techniques (eg, toys, videos, tablet computers, virtual reality devices, music, stickers, and parental presence during induction of anesthesia) may be used in place of anxiolytic medication.

The benefits of preoperative anxiolysis and sedation have to be balanced with the risk of excessive postoperative sedation [2,3] and respiratory complications, especially in children with significant OSA. Children with OSA should be monitored with continuous pulse oximetry when premedication is administered. Residual effects of the premedication may persist, especially after a relatively short surgical procedure such as tonsillectomy, and may potentially exacerbate postoperative respiratory complications and sedation, as well as prolong recovery room stay [2,3].

Analgesics — Nonopioid analgesics may be administered preoperatively or intraoperatively as part of a multimodal plan for perioperative pain management. The goal of preoperative administration is to ensure that an adequate blood level of the drug is achieved by the end of the procedure. Acetaminophen may be administered by the oral route preoperatively or by the IV or rectal route intraoperatively after induction, with the dose based on weight or age. (See "Pain in children: Approach to pain assessment and overview of management principles", section on 'Acetaminophen'.)

Nonsteroidal antiinflammatory drugs (NSAIDs) are effective at reducing the dose of opioid required for postoperative pain control, but they may increase the risk of bleeding. Therefore, NSAIDs should only be administered in consultation with the surgeon and are not usually administered preoperatively or before achieving effective hemostasis intraoperatively (see 'Nonopioid analgesics' below). We administer intravenous ketorolac intraoperatively after communicating with the ear, nose, and throat (ENT) surgeon and once hemostasis is achieved. Ketorolac can be part of opioid sparing analgesia for pediatric tonsillectomy [4].

Preoperative inhaled short-acting beta-2 agonist (SABA) — For patients who are at increased risk for postoperative respiratory complications (eg, children with moderate to severe OSA or recent upper respiratory infection (URI) [<2 weeks], or history of asthma, eczema, hay fever, or exposure to passive smoking) it is reasonable to administer a single dose of inhaled albuterol (salbutamol 200 mcg) as a preventative measure. A single dose of albuterol is generally safe and well-tolerated in children, and should be strongly considered preoperatively for children who are at risk for postoperative respiratory difficulties. The author does not routinely administer albuterol to such patients unless they are wheezing.

Besides bronchodilatory effects, albuterol, a beta₂-adrenergic agonist, inhibits the mast cell release of inflammatory mediators from mast cells, and suppresses cough [5]. Children with OSA have a high prevalence of airway inflammation, and are at higher risk for respiratory adverse events during anesthesia [6].

Pretreatment with albuterol in these high-risk children prior to tonsillectomy may reduce the likelihood of perioperative respiratory complications [7]. In a randomized placebo controlled trial of 484 children undergoing tonsillectomy (with 90 percent managed with a laryngeal mask airway [LMA]), pretreatment with albuterol reduced the frequency of perioperative laryngospasm (5 versus 12 percent), severe coughing (11 versus 33 percent), and oxygen desaturation (15 versus 23 percent), but there were similar rates of bronchospasm, stertorous airway obstruction, and stridor [7]. Albuterol's benefit was seen chiefly among children with moderate to severe OSA, who comprised 65 percent of the trial participants. In these patients, the overall frequency of any perioperative respiratory complication was 31 percent in the albuterol group compared with 58 percent in the placebo group. These rates are considerably higher than the rates reported in other studies of respiratory complications after tonsillectomy, and it is likely that many of the reported events were relatively mild and of questionable clinical significance. (see "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications", section on 'Postoperative respiratory complications'). This conclusion is supported by the finding that time in the post-anesthesia care unit (PACU) was the same in both study groups. The study did not report on the need for intervention for the respiratory events, and it is unknown whether albuterol reduces the need for intervention in PACU.

Positioning — Tonsillectomy is performed in a standard supine position with the neck extended. The head and neck should be positioned carefully for intubation and surgery. This is particularly important for children with Down syndrome, who have a high incidence of atlantoaxial instability, often without symptoms. In addition, neck extension can move the endotracheal tube and cause inadvertent extubation, especially in small children. (See "Tonsillectomy and/or adenoidectomy in children: Preoperative evaluation and care", section on 'Anatomic assessment' and "Down syndrome: Management", section on 'Atlantoaxial instability'.)

Induction of anesthesia — Inhalation induction of anesthesia is commonly performed for very young children to avoid the trauma of IV catheter placement while awake. However, IV induction is preferred for children at high risk of airway complications during induction (eg, severe OSA, uncontrolled asthma, recent upper respiratory infection, craniofacial abnormalities) (table 1).

Inhalation anesthetic agents reduce pharyngeal muscular tone [8], increase airway closing pressure [9], and cause failure of coordination of phasic activation of upper airway muscles with diaphragmatic activity [10], all of which may result in airway collapse and obstruction prior to placement of an airway device. In addition, the lower lung volumes caused by general anesthesia lead to a cephalad displacement of the mediastinum, decreasing the longitudinal tension of the upper airway and increasing susceptibility to airway collapse [11,12]. These effects are more significant in children with OSA and craniofacial abnormalities.

Laryngospasm is more common in children who undergo tonsillectomy than other procedures [13]. Prevention and management of laryngospasm are discussed separately, with management shown in an algorithm (algorithm 1). (See "General anesthesia in neonates and children: Agents and techniques", section on 'Laryngospasm'.)

IV versus inhalational induction

●Children without risk factors for airway complications – The choice between IV and inhalation induction is a balance between the desire to obtain rapid induction, rapid airway control, operating room efficiency, and the ability of the child to tolerate IV catheter placement while awake. In practice, unless there are multiple risk factors for airway complications, we usually perform an inhalational induction for children <12 years of age, and an IV induction for children ≥12 years of age and/or weight >50 kg. In older children, the decision is individualized based on the child's level of anxiety and coping. (See "General anesthesia in neonates and children: Agents and techniques", section on 'Inhalation induction'.)

●Children with OSA or other risk factors for airway complications – When possible and tolerated by the child, IV induction may be preferable for patients with risk factors for airway complications (including post-tonsillectomy bleeding) in addition to the risk associated with tonsillectomy, as shown in a table (table 1). In particular, IV induction should be performed whenever possible for children with severe OSA (see "Evaluation of suspected obstructive sleep apnea in children", section on 'Assessment of severity'). Children with severe OSA and craniofacial syndromes can quickly obstruct and desaturate. In addition to having a variety of standard laryngoscopes and blades, oral endotracheal tubes, and oral and nasal airways, anesthesia clinicians should be prepared for management of the potentially difficult airway in these patients, with difficult-airway equipment and assistance readily available (table 2).

In addition to the beneficial effects of IV induction agents in these patients, placement of an IV catheter allows immediate administration of rescue drugs during induction of anesthesia (eg, succinylcholine, atropine, epinephrine). When compared with inhalational agents, IV induction agents can more rapidly achieve a level of anesthesia deep enough for airway instrumentation. Of the IV induction agents, propofol is the agent of choice as it provides rapid onset and short duration of action, attenuates the bronchospastic response to intubation, and has antiemetic effects. (See "Anesthesia for the child with asthma or recurrent wheezing", section on 'Intravenous induction agents'.)

NMBAs — The surgeon does not require muscle relaxation for tonsillectomy. For intubation, we avoid the use of neuromuscular blocking agents (NMBAs) for most children under the age of 10 and who weigh less than 50 kg who undergo tonsillectomy, unless a rapid sequence induction and intubation is necessary (eg, for tonsillar bleeding). For children over the age of 10 who weigh more than 50 kg, we administer low-dose rocuronium, succinylcholine when necessary (eg, for rapid sequence induction and intubation), or remifentanil for intubation (see 'Remifentanil intubation' below). Cisatracurium may also be used. If nondepolarizing agents are used, adequate return of neuromuscular function should be determined before the agents are reversed and the anesthesia discontinued. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Reversal of neuromuscular block'.)

Tonsillectomy is usually a short procedure, under 30 minutes long. Thus, a short-acting NMBA is required if used to facilitate intubation. Administration of succinylcholine, a rapid onset, short-acting depolarizing NMBA, for routine intubation in children has largely been abandoned. The US Food and Drug Administration (FDA) has issued a black box warning on the use of succinylcholine in children, except for emergency airway management, over concerns for acute rhabdomyolysis and hyperkalemia with possible cardiac arrest in children with undiagnosed muscular dystrophies [14,15]. Despite this warning, succinylcholine can be used in select patients (eg, those who require rapid sequence induction), with appropriate screening prior to its administration. (See 'Strategy for anesthesia' below and "General anesthesia in neonates and children: Agents and techniques", section on 'Intravenous induction medications' and "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Succinylcholine'.)

The use of the available nondepolarizing NMBAs, which are not associated with hyperkalemia or bradycardia, is limited by their longer duration of action prior to reversal with anticholinesterase agents (45 minutes, depending on the dose). Low-dose rocuronium can be used to facilitate intubation without full paralysis in order to allow more rapid reversal. Where available, sugammadex, a chelating agent that rapidly reverses even deep levels of paralysis with rocuronium or vecuronium, allows the use of these NMBAs even for short procedures. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Sugammadex'.)

Remifentanil intubation — Remifentanil, an ultrashort-acting opioid, may be administered to facilitate intubation without the use of NMBAs for children over the age of 10 who weigh ≥50 kg. Remifentanil can provide profound suppression of airway reflexes during endotracheal intubation, without the prolonged effects of other opioids. A relatively high dose of remifentanil (3 to 4 mcg/kg IV) administered with propofol (3 to 4 mg/kg IV) and an anticholinergic (glycopyrrolate 10 mcg/kg IV or atropine 20 mcg/kg IV) has been shown to provide excellent intubating conditions in children without hemodynamic effects [16-18].

Choice of airway device — The anesthesiologist shares the airway with the surgeon during tonsillectomy. The airway device is placed in the midline and taped to the lower jaw before the mouth gag (eg, Crowe-Davis, McIvor) is placed (figure 1). The mouth gag must be positioned and opened carefully to avoid either kinking the airway device or accidental extubation.

We suggest the use of a cuffed endotracheal tube (ETT) for children undergoing adenotonsillectomy, rather than an uncuffed ETT or supraglottic airway (SGA). We find that the cuffed ETT is easily placed, allows better access to the surgical site, and provides definitive airway control that only rarely needs to be replaced. An ETT with a preformed bend (picture 1) or a wire-reinforced ETT (picture 2) should be used to avoid kinking at the mouth opening.

Advantages to a cuffed ETT include its ability to:

●Seal the trachea, preventing pulmonary aspiration of blood and secretions.

●Prevent the leakage of high-concentration oxygen into the airway, thereby reducing the chance of airway fire with the use of electrocautery at the surgical site [19].

●Prevent pollution of the operating room with volatile anesthetic gases [19].

●Provide open access to the surgical site, potentially resulting in a more effective resection [20].

Though not our usual practice, an SGA, such as an LMA, is used in some centers for tonsillectomy in children. If an LMA is used, a flexible, armored model is preferred because it is less likely to kink and has a smaller-diameter tube (and therefore can be used with the Crowe-Davis gag) compared with the standard LMA.

Potential minor benefits of an LMA include the fact that it is less stimulating to the airway and usually results in a smoother extubation with less coughing and straining. The LMA may also result in less perioperative adverse respiratory events for children with upper respiratory infections, as may be the case in children with chronic tonsillitis [21]. (See "Anesthesia for the child with a recent upper respiratory infection", section on 'Choice of airway device'.)

However, the LMA takes up more room in the patient's mouth, such that surgical exposure may be more difficult and result in less effective resection [20]. In addition, the LMA may be difficult or impossible to place or seat in patients with very large tonsils. In some cases, airway obstruction occurs when the mouth gag is opened, either because the LMA kinks or because of laryngospasm in response to the noxious stimulus of opening the mouth gag.

The literature comparing the use of an ETT with an LMA for adenotonsillectomy consists of small, retrospective studies, or those that include selected patient populations that limit generalization. A number of studies have reported rates of LMA failure during adenotonsillectomy with the need to convert to ETT, ranging from approximately 4 to 17 percent [20,22-24]. The conversion rate is lower (<1 percent) in nonobese children with normal airway anatomy and without OSA [25,26]. LMA failure may also be less likely when adenoidectomy is performed without tonsillectomy. Reasons for conversion to ETT include kinking of the LMA tubing or airway obstruction with mouth gag placement, inability to place or seat the LMA, and inadequate surgical exposure. Studies report conflicting results when comparing ETT with LMA with regard to other outcomes, including the rate of laryngospasm, coughing and gagging during surgery, operative time, and operating room time [20,22-24,27].

While the literature does not definitively support the choice of one type of airway management over the other, it does seem clear that the safe use of an LMA for adenotonsillectomy requires close coordination between the surgeon, the anesthesiologist, and the recovery room staff.

The French Association for Ambulatory Surgery (AFCA) and the French Society for Anesthesia, Intensive Care (SFAR) have published clinical practice guidelines for tonsillectomy, including guidelines for anesthesia, which agree with our preferred approach [28]. The guidelines recommend endotracheal intubation with a cuffed ETT and extubation in the presence of an anesthesiologist after complete awakening, defined as opening eyes in response to a request.

Maintenance of anesthesia — Anesthesia can be maintained using inhalation agents, most commonly sevoflurane, or by administering total intravenous anesthesia (TIVA), with infusions of hypnotic and opioid medications. We prefer inhalation agents for maintenance of anesthesia for most tonsillectomies.

For the occasional child with significant history or risk of PONV, propofol-based TIVA, is preferred, and it is usually administered along with remifentanil infusion. (See "Postoperative nausea and vomiting", section on 'Anesthetic factors' and "Maintenance of general anesthesia: Overview", section on 'Total intravenous anesthesia'.)

TIVA with propofol and remifentanil offers the following advantages for tonsillectomy, compared with other IV anesthetic agents:

●Remifentanil is an ultrashort-acting opioid that allows the maintenance of a profound level of opioid effect without the prolonged apnea seen with longer-acting opioids, which is especially important for children with severe sleep apnea.

●Propofol has an antiemetic effect, and avoidance of inhalation agents may reduce PONV [29]. (See 'Antiemetics' below.)

Whatever anesthetic agents are used for maintenance of anesthesia, adequate depth of anesthesia should be achieved prior to placement of the mouth gag by the surgeon. Opening the mouth with a gag is a sudden, noxious stimulus at the start of the procedure, which can often result in movement and even laryngospasm, breath-holding, and gagging in the absence of adequate anesthesia.

Antiemetics — Prophylaxis for PONV should be routinely administered during anesthesia for tonsillectomy. PONV occurs in 60 to over 70 percent of children who do not receive prophylactic antiemetics; the incidence of PONV can be reduced to less than 25 percent with prophylaxis [30-32]. We administer dexamethasone (0.1 to 0.5 mg/kg IV, maximum 4 mg) and a serotonergic antagonist (eg, ondansetron 0.1 mg/kg IV, maximum 4 mg) for PONV prophylaxis in children undergoing tonsillectomy [30].

Prevention of nausea and vomiting for all postoperative patients is discussed in detail separately. Studies specific to children undergoing tonsillectomy or adenotonsillectomy are reviewed here. (See "Postoperative nausea and vomiting".)

Dexamethasone — We suggest intraoperative administration of dexamethasone to prevent postoperative nausea, vomiting, and pain, as well as to reduce the time to first oral intake after tonsillectomy [33]. While a potentially increased risk of postoperative bleeding with dexamethasone is a concern, the degree of risk is unknown and is likely low. The association between dexamethasone and hemorrhage has been inconsistently observed, and the absolute risk of clinically important hemorrhage in tonsillectomy is low.

We agree with the American Academy of Otolaryngology-Head and Neck Surgery Foundation (AAO-HNSF), which recommends administration of a single intraoperative dose of IV dexamethasone in children undergoing tonsillectomy [34]. This recommendation is supported by a series of randomized trials [35-40] and 2011 and 2016 meta-analyses [33,41] that have shown that the administration of a single intraoperative dose of IV dexamethasone in children undergoing tonsillectomy results in:

●Decreased nausea and vomiting during the first 24 hours following tonsillectomy (summary odds ratio [OR] 0.23; 95% CI 0.16-0.33)

●Decreased time to first oral intake

●Decreased postoperative pain

Doses of IV dexamethasone used in these trials ranged from 0.15 to 1 mg/kg (highest maximum dose 25 mg).

Although the benefits of intraoperative dexamethasone appear to be established, the association between intraoperative dexamethasone and bleeding after tonsillectomy remains uncertain. One randomized trial was stopped because of an increase in postoperative bleeding with dexamethasone [35]. While four meta-analyses and a systematic review found no significant increase in post-tonsillectomy bleeding [33,42-45], the majority of included studies were not designed to study postoperative bleeding, and most did not include formal screening for hemorrhage. According to a 2014 meta-analysis, imprecision in the estimated odds ratio does not exclude a potential doubling of risk of hemorrhage with dexamethasone [44]. The pooled estimate of hemorrhage rate in patients treated with dexamethasone was estimated to be 6.2 percent. Dexamethasone should not be administered in the rare circumstances of tonsillectomy for suspected malignancy involving the tonsil as it may trigger cancer lysis and hence hyperkalemia.

Additional antiemetic therapy — Additional therapies to prevent PONV after tonsillectomy have been studied.

●Serotonergic antagonists – A 2006 systematic review and meta-analysis of randomized trials of antiemetic therapies found that serotonergic antagonists (eg, ondansetron, granisetron, tropisetron, dolasetron) were efficacious in preventing PONV in children undergoing tonsillectomy (summary OR for antiserotonergic agents 0.12; 95% CI 0.07-0.20) [30]. The efficacy of antiemetic therapy with serotonergic antagonists in this setting was confirmed in a 2016 meta-analysis [41]. A prophylactic dose of a serotonergic antagonist such as ondansetron can be given perioperatively in addition to intraoperative dexamethasone, particularly in patients with a history of intense PONV. Use of serotonergic antagonists for PONV is reviewed in greater detail separately. (See "Postoperative nausea and vomiting", section on 'Serotonin (5-hydroxytryptamine) receptor antagonists'.)

●Acupuncture – Limited data suggest that acupuncture may be effective for PONV prophylaxis for these patients [30,46]. Further study is required before recommending acupuncture for PONV prophylaxis.

In the setting of residual effects of anesthetics and opioids, promethazine is associated with a risk of sedation and respiratory depression and should not be used as an antiemetic in children with OSA undergoing tonsillectomy [47].

Analgesia — Tonsillectomy causes moderate to severe postoperative pain for many children, and can lead to reduced oral intake and dehydration. We suggest a multimodal approach to postoperative pain management after tonsillectomy, including the intraoperative administration of acetaminophen and dexamethasone, in order to reduce the need for opioids. This concept is particularly important for children with OSA. Children with OSA are very sensitive to the sedative and respiratory depressant effects of opioids; opioid dosing for children with OSA should be reduced by approximately 50 percent [48,49]. (See 'Sensitivity to opioids' below.)

Pain is an important cause of morbidity after tonsillectomy and can result in dysphagia, decreased oral intake, dehydration, and weight loss. Pain is worst in the first few postoperative days but significant pain may last up to two weeks [50].

The effects of the specific surgical technique and patient risk factors associated with increased postoperative pain are discussed separately (see "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications", section on 'Pain' and "Tonsillectomy in adults: Techniques and perioperative issues", section on 'Techniques'). Neither of these variables should be used to modify perioperative opioid doses.

Nonopioid analgesics — Multimodal analgesics should routinely be scheduled for postoperative throat pain and otalgia [34]. For most patients, we administer regularly scheduled postoperative acetaminophen, ibuprofen, dexamethasone, and low doses of opioid on as-needed basis only. We do not routinely administer NSAIDs intraoperatively due to a higher risk of bleeding following tonsillectomy.

●Acetaminophen – Acetaminophen may be administered by oral route (15 mg/kg preoperatively), by rectal suppository (40 mg/kg as loading dose) after induction, or by IV bolus (15 mg/kg). IV acetaminophen has been shown to reduce the need for rescue analgesics after tonsillectomy and to slightly reduce recovery room stay [51-53]. While the rectal route of administration may result in an unpredictable plasma concentration [54,55], the duration of analgesia with acetaminophen may be longer with rectal administration (10 hours with 40 mg/kg rectally) compared with IV administration (7 hours with 15 mg/kg IV) for pain after adenotonsillectomy [56].

●NSAIDs – We do not routinely administer IV NSAIDs during tonsillectomy because of the risk of tonsillar bleeding. However, the author administers IV ketorolac in the operating room immediately after tonsillectomy if the surgeon is comfortable with hemostasis. In our experience, ketorolac administered this way has provided effective opioid-sparing analgesia without an increase in post-tonsillectomy bleeding [57]. In clinical trials, NSAIDs have been shown to be as effective in treating post-tonsillectomy pain as opioid medications, with a reduction in PONV, sedation, and respiratory depression [58-61].

Some studies have reported increased post-tonsillectomy bleeding with preoperative [62] or intraoperative ketorolac administration [61,63,64], while others have reported no difference [65,66]. A systematic review of trials on the effects of NSAIDs on bleeding after pediatric tonsillectomy included 15 studies with over 1100 children [67]. The authors concluded that there was insufficient evidence to exclude an increase in postoperative bleeding with ketorolac administration.

Small studies suggest that the selective COX-2 inhibitors parecoxib [68,69] and celecoxib [70] (which have no effects of platelet function at clinically used doses) can reduce post-tonsillectomy pain, though further study is required to evaluate bleeding and cardiovascular risks. The appropriate dose of parecoxib is 0.65 to 0.9 mg/kg IV in children 2 to 12 years [68,69]. Doses >1 mg/kg provide no additional benefit over lower doses. Parecoxib is not available in the United States.

Where available, rectal diclofenac (1 mg/kg) has been administered after induction of anesthesia as part of a regimen for postoperative pain control after tonsillectomy.

●Dexamethasone – Dexamethasone reduces pain and swelling, in addition to reducing PONV. Intraoperative administration of dexamethasone should be part of the multimodal plan for management of pain after tonsillectomy, except for cases of suspected tonsillar malignancy. (See 'Dexamethasone' above.)

●Dexmedetomidine – We often administer dexmedetomidine 0.5 mcg/kg IV during anesthesia as an opioid sparing analgesic, particularly in children with preoperative anxiety or prior history of emergence delirium and/or complications from opioids (eg, PONV). We administer 0.5 mcg/kg IV, as higher doses (>1 mcg/kg) can cause sedation and prolonged stay in the post-anesthesia care unit (PACU).

•Analgesic effects – Dexmedetomidine is a selective alpha adrenoreceptor agonist with analgesic and sedative properties, without respiratory depressant effects [71-74], that has been shown to reduce emergence delirium in children [75-77]. Dexmedetomidine is often used as part of multimodal, opioid-sparing anesthesia for a variety of surgical procedures, but studies evaluating dexmedetomidine administered before and during tonsillectomy have reported conflicting results. Some have reported reduced postoperative opioid requirements when compared with fentanyl [72,78], while others have shown no opioid-sparing effect when dexmedetomidine was compared with morphine [71,73].

In a retrospective cohort study of 941 children undergoing adenotonsillectomy, intraoperative dexmedetomidine was associated with a dose dependent reduction in perioperative opioid consumption, but a small increase in time in the PACU [79]. Administration of intraoperative dexmedetomidine 0.5 mcg/kg IV was associated with reduced perioperative oral morphine equivalent use of approximately 0.1 mg/kg, and an increase of >5 minutes in PACU stay. In addition, perioperative opioid consumption was similar in patients with and without obstructive sleep apnea.

•Facilitating deep extubation – In addition to possible analgesic effects, premedication with dexmedetomidine may facilitate deep extubation and postoperative recovery after tonsillectomy, if deep extubation is planned [80] (see 'Emergence from anesthesia' below). In one study, children undergoing tonsillectomy were randomly assigned to receive preoperative dexmedetomidine 1 or 2 mcg/kg IV, along with low-dose sevoflurane prior to deep extubation at the end of surgery, or no dexmedetomidine and high-dose sevoflurane prior to deep extubation [80]. Children who received dexmedetomidine at either dose required less postoperative opioid, had less postoperative delirium, and required insertion of an oral airway less often, than children who received no dexmedetomidine. Length of stay in the post-anesthesia care unit was shortest in children who received dexmedetomidine 1 mcg/kg IV. At higher doses (eg, 2 to 4 mcg/kg IV after induction of anesthesia), dexmedetomidine may prolong time to extubation and recovery room stay [78]. Further study is required before recommending routine perioperative administration of dexmedetomidine for tonsillectomy.

Emergence delirium in children is discussed separately. (See "Emergence delirium and agitation in children".)

●Ketamine – Ketamine is an N-methyl-D-aspartate (NMDA) antagonist that has an opioid-sparing effect when administered at low doses during anesthesia in adults [81], though its efficacy in children is controversial [82]. Meta-analyses of literature on ketamine for postoperative pain in children have reported decreased recovery-room pain intensity without side effects, but no postoperative opioid-sparing effects [83,84]. We avoid administration of ketamine for tonsillectomy because it may increase postoperative agitation and secretions without providing benefit.

Sensitivity to opioids — The aim of multimodal pain control is to provide adequate analgesia while reducing the need for postoperative opioids. A short-acting opioid (eg, fentanyl 1 mcg/kg IV) is commonly administered during or immediately after induction of anesthesia for tonsillectomy. High doses of longer-acting opioid should be avoided, if possible, especially in children with OSA, in order to reduce the risk of postoperative sedation and respiratory depression.

Children with OSA are more susceptible to the respiratory depressant and airway effects of opioids compared with those without OSA; opioid dosing for children with OSA should be reduced by approximately 50 percent [49,85]. Small blinded trials and observational studies have reported that sensitivity to opioids in children with OSA correlates with the severity of recurrent hypoxemia. Children with severe OSA who desaturate to ≤85 percent during a sleep study have been shown to require approximately 50 percent less morphine for pain after tonsillectomy compared with those whose oxygen saturation nadir was >85 percent [48,86,87]. Central and peripheral mechanisms partially explain this increased sensitivity, which may be related to upregulation of mu-opioid receptors caused by hypoxia [88]. Certain genotypes may increase sensitivity to opioids and the risk of respiratory depression and PONV in children undergoing tonsillectomy [89-93].

Because of the risk of postoperative respiratory depression and airway obstruction, the lowest effective doses of short-acting opioids should be administered during anesthesia for tonsillectomy, in combination with nonopioid analgesics (eg, acetaminophen and dexamethasone).

Local anesthesia — The use of local anesthesia during tonsillectomy is controversial. A variety of medications, alone and in combination, have been injected into the tonsillar bed in an effort to provide postoperative analgesia, with variable success. In a systematic review of six small randomized trials involving adults or children (mostly adults) who underwent tonsillectomy, injection or topical application of local anesthetics was not found to reduce pain or the need for supplemental analgesics [94]. In addition, there are reports of significant complications related to the use of local infiltration after tonsillectomy, including ventricular arrhythmias, infection, airway compromise, and cardiac arrest [95-97].

Music therapy — In one randomized clinical trial, children undergoing adenotonsillectomy who received intraoperative auditory stimulation with music demonstrated a clinically meaningful decrease in severe pain on awakening (0.63, 98% CI 0.43-0.84) and emergence delirium (0.47, 98% CI 0.21-0.75) [98]. Further research is required to validate the effectiveness and opioid-sparing effects of music therapy in children.

Antimicrobials — The use of antimicrobial prophylaxis at the time of, or for short periods following, tonsillectomy does not improve postoperative outcomes. This issue is discussed in detail separately. (See "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications", section on 'No role for antimicrobial prophylaxis'.)

Emergence from anesthesia — Emergence from anesthesia after tonsillectomy can be a challenge as children who undergo tonsillectomy are at high risk of laryngospasm and increased airway reactivity. To minimize the possibility of laryngospasm, the oropharynx should be thoroughly suctioned prior to emergence to remove blood and secretions, and the stomach emptied of blood, with an orogastric tube. All children should be extubated and positioned for recovery in the lateral position with the neck slightly extended (the "tonsil position") to allow secretions to drain away from the oropharynx. All children should be cared for until wide awake by clinicians who are experienced in basic airway support, such as jaw thrust and oral or nasal airway placement.

We extubate most patients after tonsillectomy awake. The choice between awake and deep extubation is based on patient factors and clinician preference. Awake extubation allows the return of airway tone and airway protective reflexes, and theoretically reduces the risk of post-extubation airway obstruction and laryngospasm. However, in children with risk factors for bronchospasm, or to minimize coughing for surgical reasons, deep extubation may be preferred.

In patients with risk factors for difficulty with airway management (eg, craniofacial anomalies) awake extubation is preferred.

In particular, children with severe OSA should be extubated awake with full recovery of strength and minimal residual anesthetic effect. Even under these conditions, significant airway obstruction and respiratory complications may occur. Airway manipulation and support, with jaw thrust, oral, or nasal airway placement, may be necessary. (See "General anesthesia in neonates and children: Agents and techniques", section on 'Deep versus awake extubation'.)

SGAs are most often removed while the child is still deeply anesthetized. Awake removal of an SGA may be preferred in patients who have copious upper airway secretions, blood in the airway, OSA, or anatomic airway obstruction. (See "General anesthesia in neonates and children: Agents and techniques", section on 'Deep versus awake removal of SGA'.)

The child should be transported to the recovery area while breathing supplemental oxygen by facemask or nasal cannula.

POSTOPERATIVE CARE — While most children who undergo tonsillectomy have a smooth postoperative course, significant complications can occur. Post tonsillectomy hemorrhage is an infrequent but potentially life-threatening complication, and is discussed below (see 'Post-tonsillectomy hemorrhage' below and "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications", section on 'Hemorrhage'). Children with sleep-disordered breathing and obstructive sleep apnea (OSA) are at increased risk for death and/or permanent neurologic injury with apneic episodes in the postoperative period [99,100].

Most tonsillectomies (with or without adenoidectomy) are performed in an outpatient setting, and children are discharged a few hours after surgery. However, children who are at higher risk for complications (eg, age <3 years, those with Down syndrome or other comorbidities such as neuromuscular disorders and airway anomalies, and/or a history of severe OSA) may be admitted overnight for observation after tonsillectomy. Increased risk in small children is supported by a retrospective single-center study including 805 children who underwent tonsillectomy between 2012 and 2014 [101]. Weight <18 kg was an independent risk factor for postoperative respiratory failure, as indicated by unplanned transfer to intensive care and the need for supported ventilation in the first 60 minutes in the post-anesthesia care unit. The appropriate setting for postoperative monitoring and indications for overnight hospitalization are discussed separately. (See "Adenotonsillectomy for obstructive sleep apnea in children", section on 'Operative setting'.)

It is important to differentiate emergence delirium from pain and other etiologies for postoperative agitation in the immediate post-anesthesia period to guide appropriate treatment (table 3). Emergence delirium in children is discussed separately. (See "Emergence delirium and agitation in children".)

General postoperative care, including disposition and pain management, and potential complications in children who have undergone tonsillectomy (with or without adenoidectomy) are reviewed in more detail separately. (See "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications".)

POST-TONSILLECTOMY HEMORRHAGE — Post-tonsillectomy bleeding is an infrequent, potentially life-threatening complication of tonsillectomy. Bleeding may occur in the first 24 hours after surgery (primary hemorrhage) or may be delayed, most commonly occurring 5 to 12 days after surgery (secondary hemorrhage). (See "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications", section on 'Hemorrhage'.)

Challenges for the anesthesiologist — Post-tonsillectomy hemorrhage is a surgical emergency that presents significant challenges for the anesthesiologist. In particular, the child with a bleeding tonsil may be severely hypovolemic, and if hypovolemia is unrecognized or uncorrected, the child may become profoundly hypotensive with induction of anesthesia.

Anesthetic concerns include the following:

●Hypovolemia – Primary concerns prior to induction of anesthesia are the identification and correction of hypovolemia related to tonsillar bleeding. It can be difficult or impossible to quantify the blood loss from hemorrhage after tonsillectomy. Much of the blood can be swallowed, and vomited blood may be very difficult to estimate. Dizziness and orthostatic vital signs suggest hypovolemia. Hypovolemic patients are at high risk of hypotension with induction of anesthesia.

●Full stomach – Most blood is swallowed, such that these patients have full stomachs and are at risk of aspiration.

●Difficult intubation – Even if intubation for the original procedure went smoothly, endotracheal intubation in the setting of post-tonsillectomy bleeding may be complicated due to tissue swelling and blood in the oropharynx.

●Anemia – Depending on the degree of blood loss and the amount of crystalloid replacement, the patient may present with significant anemia.

A retrospective study of 475 anesthetics for bleeding after tonsillectomy reported that hypoxemia was the most common adverse event [102]. Hypoxemia (oxyhemoglobin saturation <90 percent) occurred in approximately 10 percent of cases. Twice as many episodes of hypoxemia occurred during extubation or emergence than did during induction. Hypotension occurred in approximately 2.5 percent of cases, and 2.7 percent of patients were difficult to intubate, none of whom were difficult to intubate during tonsillectomy.

Strategy for anesthesia — We suggest the following plan for anesthesia for the bleeding tonsil:

●Preoperative evaluation – When possible, a quick review of the prior anesthesia record, including details of airway management, orthostatic vital signs, and review of medical history, should be performed.

●Preparation – Additional anesthesia personnel should be summoned to assist whenever possible. Laryngoscopes; styletted, cuffed endotracheal tubes (ETTs) in various sizes; and a suction apparatus should be prepared in duplicate, as blood can block tubing or the laryngoscope's light source. Routine monitors should be applied, including noninvasive blood pressure, electrocardiogram, and pulse oximetry.

●Intravenous (IV) access – IV access must be established prior to induction of anesthesia, and, if there is any indication of hypovolemia, vigorous volume resuscitation with crystalloid and/or colloid solutions or blood should be performed prior to induction. Induction should not be performed until resuscitation is complete.

●Positioning – Preoxygenation should be performed with the patient in the lateral decubitus position, head down, in order to allow drainage of blood away from the pharynx. The child should be turned supine as rapid sequence induction and intubation (RSII) is performed.

●Rapid sequence induction and intubation (RSII) – RSII should be performed. In this setting, even with apparently adequate resuscitation, a reduced dose of propofol (1 to 2 mg/kg IV), ketamine (1 to 2 mg/kg IV), or etomidate (0.2 mg/kg IV) should be administered for induction, immediately followed by the combination of atropine (0.02 mg/kg IV) plus succinylcholine (1.5 to 2 mg/kg IV) or rocuronium (1.2 mg/kg IV). Cricoid pressure should be applied, and endotracheal intubation should be performed as quickly as possible. (See "Rapid sequence induction and intubation (RSII) for anesthesia".)

●Maintenance of anesthesia – These surgical procedures are not painful and may be very brief if control of a bleeding blood vessel is easily accomplished, though in cases of coagulopathy, hemostasis may not be easily accomplished. Maintenance of anesthesia should aim for rapid emergence and recovery of airway protective reflexes, with minimal, if any, administration of long-acting opioids.

●Emergence from anesthesia – The patient must be fully awake, with neuromuscular blocking agent (NMBA) reversed, and able to protect the airway prior to extubation from anesthesia for tonsillar bleeding. The oropharynx should be carefully suctioned and the stomach emptied with an orogastric tube, realizing that clotted blood may not be effectively removed. The patient should be extubated in the lateral decubitus or seated upright position and transported to the recovery room while breathing supplemental oxygen.

ADENOIDECTOMY — Adenoidectomy is a common pediatric surgical procedure that is performed alone or in conjunction with tonsillectomy. Patients usually go home the day of surgery, except for patients with obstructive sleep apnea (OSA) and those with comorbid conditions that increase the risk of postoperative respiratory adverse events (eg, Down syndrome, neuromuscular disorders, syndromes with craniofacial abnormalities). (See "Tonsillectomy (with or without adenoidectomy) in children: Postoperative care and complications".)

Anesthesia concerns are similar to those for tonsillectomy. Compared with tonsillectomy, adenoidectomy alone is a shorter, less painful procedure with usually more straightforward airway management. Adenoidectomy may be performed using a laryngeal mask airway (LMA). Postoperative pain is usually well controlled with as-needed acetaminophen and nonsteroidal antiinflammatory drugs (NSAIDs), typically without the need for opioids. (See "Adenoidectomy in children: Postoperative care and complications", section on 'General postoperative care'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pediatric anesthesia" and "Society guideline links: Tonsillectomy and adenoidectomy in children".)

SUMMARY AND RECOMMENDATIONS

●Preanesthesia evaluation – The most common indication for tonsillectomy is airway obstruction from sleep-disordered breathing or obstructive sleep apnea (OSA). Preoperative evaluation should include screening for OSA, as well as a sleep study if clinical manifestations of OSA are present. Children with OSA are at increased risk of airway obstruction, laryngospasm, and bronchospasm during anesthesia, and they are at increased risk of adverse respiratory events after surgery. (See 'Preanesthetic evaluation' above.)

●Induction of anesthesia – Our approach to induction technique largely depends on the ability of a child to tolerate placement of an intravenous (IV) catheter while awake. For young children (≤6 years), in practice, this typically results in performance of an inhalational induction. For older children (≥12 years) and for children of any age with severe OSA, we suggest IV rather than inhalational induction (Grade 2C). (See 'Induction of anesthesia' above.)

●Airway management – We suggest the use of a cuffed endotracheal tube (ETT) for airway management, rather than an uncuffed ETT or a laryngeal mask airway (LMA), for tonsillectomy (Grade 2C). If an LMA is used, close coordination between the surgeon, the anesthesiologist, and the recovery room staff is required. (See 'Choice of airway device' above.)

●Maintenance of anesthesia

•Opioid doses – Children with OSA are more sensitive to opioids and sedatives than children without OSA. For children with severe OSA, we recommend using reduced doses of intraoperative opioids (Grade 1B). We typically reduce opioid doses by 50 percent. The lowest effective doses of short-acting opioids should be administered during anesthesia for tonsillectomy and after surgery, in combination with nonopioid analgesics (eg, acetaminophen and dexamethasone). (See 'Analgesia' above and 'Sensitivity to opioids' above.)

•Intraoperative dexamethasone – We suggest a single intraoperative dose of dexamethasone in children undergoing tonsillectomy with or without adenoidectomy (Grade 2B). An intraoperative dose of IV dexamethasone is effective in decreasing postoperative nausea, vomiting, pain, and time to first oral intake. We also administer a serotonergic antagonist (eg, ondansetron 0.1 mg/kg IV) for prophylaxis for postoperative nausea and vomiting (PONV). (See 'Antiemetics' above.)

●Emergence – For emergence, the oropharynx should be thoroughly suctioned, and the stomach emptied, with an orogastric tube. Most patients, particularly those with OSA, should be extubated awake and able to protect the airway. (See 'Emergence from anesthesia' above.)

●Postoperative care

•Pain control – We use a multimodal approach to postoperative pain control for tonsillectomy, beginning with intraoperative administration of a short-acting opioid (eg, fentanyl 1 mcg/kg IV), acetaminophen (15 mg/kg IV or 40 mg/kg rectally), and dexamethasone (0.1 to 0.5 mg/kg IV). We administer regularly scheduled acetaminophen, ibuprofen, and dexamethasone postoperatively, and add low dose opioids post operatively as needed. (See 'Analgesia' above.)

•Disposition – Children at particular risk for postoperative respiratory events should be admitted to the hospital after tonsillectomy for observation, including those under three years of age, those with Down syndrome or other comorbidities such as neuromuscular disorders and airway anomalies, and those with severe OSA. (See 'Postoperative care' above.)

●Postoperative hemorrhage – Post-tonsillectomy hemorrhage is a surgical emergency. Patients may present with significant hypovolemia, anemia, a full stomach, and difficult airway management. Components of anesthetic management should include (see 'Strategy for anesthesia' above):

•Preparation for difficult airway management

•IV fluid resuscitation prior to induction

•Rapid sequence induction with either low-dose propofol, ketamine, or etomidate, followed by placement of a cuffed ETT

•Emergence when fully awake