Overview
Tonsillectomy is one of the most frequently performed surgical procedures, yet pain management has historically been suboptimal — partly due to longstanding (and largely unsupported) concerns that NSAIDs and steroids might increase post-tonsillectomy haemorrhage risk. PROSPECT 2021 represents a significant step toward evidence-based clarity on these long-debated safety questions.
PROSPECT 2021 summary (Aldamluji et al.): The analgesic regimen should include paracetamol, NSAIDs, and IV dexamethasone, with opioids as rescue. Notably, acupuncture and postoperative honey are also specifically recommended adjuncts — an unusual but evidence-supported addition compared to most other PROSPECT guidelines. Ketamine (children only), dexmedetomidine, or gabapentinoids may be considered second-line when first-line agents are contraindicated.
NSAIDs and Bleeding Risk — Setting the Record Straight
The historical concern that NSAIDs increase post-tonsillectomy haemorrhage (PTH) risk is not supported by the current evidence base. Multiple meta-analyses, including two from 2013 and the underlying PROSPECT review, found no increased risk of bleeding with NSAID use. This represents an important practice point: NSAIDs should not be withheld from tonsillectomy patients on this historical basis alone.
This matters clinically because avoiding NSAIDs unnecessarily removes one of the most effective non-opioid analgesics from the regimen, pushing patients toward greater opioid reliance — which carries its own well-documented risks, particularly in children (see codeine warning below).
Analgesic Pathway
- Paracetamol PO (weight-based in children)
- NSAID PO (e.g. ibuprofen) — no contraindication from bleeding-risk concern
- Dexamethasone IV — see dosing below
- Paracetamol IV (if not given pre-op)
- Intra-operative acupuncture — PROSPECT-supported adjunct
- Ketamine — children only, second-line
- Paracetamol regular, weight-based dosing in children
- NSAID regular (e.g. ibuprofen)
- Honey — oral, post-operative
- Postoperative acupuncture — where service available
- Opioid PRN rescue — avoid codeine in children (see warning)
✓ Recommended | ◎ Optional/conditional. Based on PROSPECT 2021 (Aldamluji et al.).
Systemic Analgesia
| Agent | Adult Dose | Paediatric Dose | Notes |
|---|---|---|---|
| Paracetamol | 1 g PO/IV QDS | 15 mg/kg QDS (max 60 mg/kg/day) | Regular throughout — foundational |
| Ibuprofen | 400 mg PO TDS | 10 mg/kg TDS | No increased bleeding risk per current evidence |
| Dexamethasone | 8 mg+ IV single dose | ≥0.15 mg/kg IV single dose | See dedicated section below |
| Oxycodone | 5–10 mg PO PRN | Weight-based, specialist-guided PRN | Rescue only; preferred over codeine in children |
Codeine is contraindicated in children following tonsillectomy/adenotonsillectomy for obstructive sleep apnoea, and used with caution more broadly in children, due to variable CYP2D6 metabolism — ultra-rapid metabolisers can develop life-threatening respiratory depression from standard doses, with documented paediatric deaths reported following tonsillectomy. Regulatory agencies (FDA, MHRA) have issued specific warnings. Use weight-appropriate paracetamol/NSAID combinations and, where opioid rescue is needed, oxycodone or morphine under appropriate monitoring instead.
Dexamethasone — Dose & Bleeding Risk
Dexamethasone consistently reduced postoperative pain across the reviewed studies, with no consistent dose-response relationship identified. Paediatric studies showing benefit used doses of at least 0.15 mg/kg; adult studies used 8 mg or more. PROSPECT found no evidence of increased bleeding risk with dexamethasone or other glucocorticoid-related side effects, though acknowledged that none of the included studies had systematically/primarily investigated this safety outcome.
| Population | Dose | Notes |
|---|---|---|
| Adults | 8 mg IV (or more) | Single intraoperative dose |
| Children | ≥0.15 mg/kg IV | Single intraoperative dose; max per local paediatric protocol |
Nuance on the bleeding question: While PROSPECT and most meta-analyses found no overall increase in bleeding rate with dexamethasone, one meta-analysis identified a small increase in the rate of operative intervention for bleeding specifically (3% vs 1.5%) without a corresponding increase in overall bleeding incidence — and this finding has not been consistently replicated. The overall evidence still supports dexamethasone's net benefit (analgesia + reliable PONV reduction), but this nuance is worth being aware of rather than treating the safety question as entirely closed.
Honey & Acupuncture — Evidence-Supported Adjuncts
Postoperative honey and both intra-operative and postoperative acupuncture are specifically recommended by PROSPECT for tonsillectomy — a distinctive feature of this guideline compared to most other procedures on this site, where such adjuncts are typically positioned as low-evidence last resorts rather than core recommendations.
Honey is typically given orally in the postoperative period (e.g. a teaspoon several times daily) — proposed mechanisms include local anti-inflammatory and wound-healing effects on the tonsillar fossa mucosa, alongside a soothing/demulcent effect. This is a low-cost, low-risk, well-tolerated adjunct worth incorporating into routine post-tonsillectomy discharge advice in patients without honey allergy/intolerance and old enough to safely consume it (avoid in infants under 12 months — infant botulism risk, though not relevant to typical tonsillectomy age range).
Second-Line Options
When first-line agents (paracetamol, NSAIDs, dexamethasone) are contraindicated or insufficient, PROSPECT identifies the following as reasonable second-line considerations:
| Agent | Population | Notes |
|---|---|---|
| Ketamine | Children only | Evidence specifically assessed in paediatric studies; sub-anaesthetic dosing |
| Dexmedetomidine | Adults and children | IV infusion; haemodynamic monitoring required |
| Gabapentinoids | Adults primarily | Weak-to-moderate effect demonstrated; balance against sedation risk |
Inconsistent evidence was found for local anaesthetic infiltration, antibiotics, and magnesium sulfate; limited evidence for clonidine — none of these are core recommendations.
Local Anaesthetic Infiltration — Important Safety Warning
A study investigating bilateral local anaesthetic injection (bupivacaine 0.25–0.5%, 3–10 mL) into the lateral pharyngeal space was terminated early after 2 of the first 4 children enrolled developed severe upper airway obstruction following tracheal extubation. This is a serious, specific safety signal — bilateral peritonsillar/lateral pharyngeal LA infiltration is not a benign intervention and should be approached with significant caution, particularly in children, given this documented risk of acute postoperative airway compromise.
Topical local anaesthetic application (e.g. levobupivacaine-soaked swabs, or injection directly into the tonsillar fossa rather than the lateral pharyngeal space specifically) is generally considered a lower-risk alternative if local anaesthetic technique is desired, though overall PROSPECT evidence for LA infiltration techniques remains inconsistent regarding analgesic benefit.
Practical Analgesic Protocol
Suggested protocol for routine elective tonsillectomy:
References
- Aldamluji N, Burgess A, Pogatzki-Zahn E, Raeder J, Beloeil H; PROSPECT Working Group. PROSPECT guideline for tonsillectomy: systematic review and procedure-specific postoperative pain management recommendations. Anaesthesia. 2021;76(7):947–961. PMC8247026 ↗
- Dexamethasone and postoperative bleeding after tonsillectomy and adenotonsillectomy in children: A meta-analysis of prospective studies. PubMed. PMID 22447657 ↗
- Jensen DR. Pharmacologic management of post-tonsillectomy pain in children. World J Otorhinolaryngol Head Neck Surg. 2021;7(3):186–193. PMC8356107 ↗
- Systematic review and meta-analysis of pain management after tonsillectomy. Sci Rep. 2024. Nature ↗
- Systematic review of analgesics and dexamethasone for post-tonsillectomy pain in adults. Br J Anaesth. 2019. ScienceDirect ↗
- FDA Drug Safety Communication: codeine use in certain children after tonsillectomy and/or adenoidectomy.