Reimer 1998.

Methods	Randomized, controlled, double‐blinded study
Participants	ASA I‐II children from 7‐12 years old, undergoing adenotonsillectomy. 41 enrolled but only 36 in final analysis. Removals due to breach of protocol (3); inability to cooperate with VAS due to ADD (1) and bronchospasm on extubation requiring adrenaline nebuliser (1) Exclusions: inability to understand English, contraindications to any of the medications in the study, obesity (weight > 90th percentile by nomogram), inability to use a visual analogue scale, use of any preoperative sedative, hypnotic or analgesic medications
Interventions	Clonidine group: clonidine 4 μg/kg orally to maximum 200 μg, 60‐90 min preoperatively; then intravenous placebo immediately post‐induction Fentanyl group: placebo orally, 60‐90 min preoperatively; then intravenous fentanyl 3 μg/kg immediately post‐induction
Outcomes	Patients receiving morphine (0.05 mg/kg) in PACU Total morphine given Number of morphine doses given Number receiving codeine and/or paracetamol (acetaminophen) in day care unit Vomiting within 24 hours Use of analgesia post‐discharge Excessive sedation Also measured preoperative anxiety and sedation scores, satisfaction scores, and vital signs
Notes	Local anaesthetic infiltration of tonsillar bed by surgeon prior to incision
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random numbers, done by pharmacy
Allocation concealment (selection bias)	Low risk	Sealed envelopes. Randomization code kept by pharmacy; also held in sealed envelopes by one investigator in case the information was needed after hours
Blinding (performance bias and detection bias) All outcomes	Low risk	All parties, including anaesthetist, patient, and observer blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	All patients included in the analysis
Selective reporting (reporting bias)	Unclear risk	None evident
Other bias	Low risk	None apparent; patient demographics and surgery type distributed evenly between groups