Preventive effect of ramelteon on emergence agitation after general anaesthesia in paediatric patients undergoing tonsillectomy: a randomised, placebo-controlled clinical trial

Emergence agitation is a common and clinically significant complication after general anaesthesia in children, characterized by crying, excitation, thrashing, disorientation, and incoherence. It increases the risk of self-injury, accidental removal of dressings and lines, and often necessitates additional sedatives or analgesics, which may cause respiratory depression. The risk of emergence agitation is particularly high after tonsillectomy in young children, making preventive strategies important in this population. Pharmacologic options such as midazolam, fentanyl, and propofol have been used but carry risks of respiratory depression. Melatonin, a pineal hormone regulating sleep-wake rhythm, has been suggested by a systematic review and meta-analysis to reduce emergence agitation without respiratory depression in children, though confidence intervals were wide and further trials were recommended. Ramelteon, a melatonin receptor agonist with higher affinity for MT1 and MT2 receptors than melatonin, has shown preventive effects on delirium in adults in case series, retrospective studies, and randomized trials. However, its role in preventing paediatric emergence agitation has not been evaluated. This trial aimed to test the hypothesis that preoperative ramelteon prevents emergence agitation in children undergoing tonsillectomy under general anaesthesia.

- Emergence agitation occurs frequently in children post-anaesthesia, especially after tonsillectomy, and can lead to adverse events and need for additional medications that may depress respiration. - Prior pharmacologic approaches (midazolam, fentanyl, propofol) can mitigate agitation but are associated with respiratory depression risk. - A systematic review/meta-analysis suggested melatonin (0.1–0.5 mg·kg⁻¹) can prevent emergence agitation in children, with a stronger effect at higher doses, but with wide confidence intervals necessitating further research. - Ramelteon is a selective melatonin receptor agonist (higher affinity than melatonin for MT1/MT2) and has been associated with delirium prevention in adults (case reports, retrospective studies, and randomized controlled trials), but had not been studied for paediatric emergence agitation prior to this trial.

Design and oversight: Prospective, double-blind, parallel-group, placebo-controlled, randomized clinical trial at Kanagawa Children’s Medical Center, Yokohama, Japan. IRB approval (No. 98-03) on 20 September 2016; registered in UMIN (UMIN000024379, 12 October 2016). Conducted per Declaration of Helsinki and reported per CONSORT. Independent oversight by Mirai Clinical Trial Support Centre and the hospital’s clinical trial supporting team. Written informed consent was obtained from parents. Participants: Children aged 18–119 months with ASA-PS 1–2 scheduled for tonsillectomy under general anaesthesia (October 2016–September 2018). Exclusions: history of mental retardation/disease, psychotropic or anticonvulsant use, cardiac disease, cerebral surgery, or same-day admission preventing adequate premedication interval. Randomization and blinding: Computer-generated allocation (randomization.com). Only hospital pharmacists knew allocations; patients, parents, anaesthesiologists, outcome assessors, and data collectors were blinded. Allocation confirmed postoperatively by a pharmacist. Interventions: Ramelteon group received 0.1 mg·kg⁻¹ ramelteon dissolved in 5 mL lactose-containing syrup before induction. Placebo group received 5 mL of identical syrup without active drug. Anaesthesia protocol: Preoperative fasting per standard. Induction via inhaled 8% sevoflurane with nitrous oxide and oxygen via face mask; after loss of consciousness, nitrous oxide discontinued, sevoflurane reduced to 5% in 100% oxygen, IV access obtained. Intraoperative fentanyl 2 µg·kg⁻¹ IV; additional fentanyl 1 µg·kg⁻¹ if BP/HR increased >10% from baseline; rocuronium as needed. Maintenance with sevoflurane 2–3% in 100% oxygen. Paracetamol suppository 30 mg·kg⁻¹ pre-incision. Extubation after adequate spontaneous breathing in 100% oxygen. Rescue for emergence agitation at clinician discretion: fentanyl, propofol, or dexmedetomidine. Outcomes and measurements: Pre-induction behaviour/anxiety assessed by Paediatric Anaesthesia Behaviour (PAB) scale. In PACU, every 5 min assessments with Paediatric Anaesthesia Emergence Delirium (PAED) scale and Children’s Hospital Eastern Ontario Pain Scale (CHEOPS); worst (peak) scores recorded. Pulse rate and oxygen saturation monitored; postoperative vomiting (POV) assessed over 24 h. Primary outcome: incidence of emergence agitation defined as PAED ≥10. Secondary outcomes: PAED scores, Aono’s scores, incidence of POV, CHEOPS scores, number requiring rescue drugs for agitation, PAB scores, incidence of desaturation (<95% in room air) pre-anaesthesia and in PACU, and time to recovery from anaesthesia in PACU. Sample size and statistical analysis: Assumed agitation incidence 67% (placebo) vs 21% (ramelteon) based on prior data. With alpha 0.05 and power 0.80, 22 per group required; target enrollment 50 (25/group) allowing for dropouts. Analyses on modified intention-to-treat (all randomized with outcome data). Normality tested by Shapiro–Wilk. Normally distributed continuous variables: mean±SD, unpaired t-test. Non-normal: median [IQR], Mann–Whitney U. Categorical: number (%), Fisher’s exact test. Reported mean/median differences or risk ratios with 95% CIs; median difference CIs via bootstrap (2000 resamples). Two-sided P<0.05 significant. Analyses in R 3.4.3.

- Enrollment and analysis: 50 randomized; 2 excluded post-randomization (surgery cancellation; missing primary outcome due to return to OR), yielding 48 in modified ITT. Three patients did not receive premedication but were included in analysis. - Primary outcome: Incidence of emergence agitation identical between groups: 16/24 (66.7%) placebo vs 16/24 (66.7%) ramelteon; risk ratio 1.00 (95% CI 0.67–1.49); P=0.99. - Secondary outcomes (all non-significant between groups): • PAED peak score: placebo 13.5 [7.5–18.0] vs ramelteon 13.5 [6.8–19.2]; median difference 0.0 (95% CI −4.8 to 5.6); P=0.79. • Aono’s score: 3 [1–4] vs 3 [1.75–4]; median difference 0.0 (95% CI −1.1 to 1.4); P=0.80. • Rescue drug use for agitation: 45.8% vs 54.2%; RR 1.18 (95% CI 0.67–2.09); P=0.77. • Discharged in PACU: 25.0% vs 4.2%; RR 0.17 (95% CI 0.02–1.28); P=0.10. • Post-operative vomiting within 24 h: 25.0% vs 20.8%; RR 0.83 (95% CI 0.29–2.37); P=0.99. • Time to recovery from anaesthesia (min): 16 [12–22] vs 12 [7–18]; median difference −4.0 (95% CI −9.1 to 0.1); P=0.08. • CHEOPS pain score: 12.0 [7.5–13.0] vs 11.5 [9–12]; median difference −0.5 (95% CI −2.5 to 1.6); P=0.56. • PAB score: [1–2] vs [1–2]; median difference 0.0 (95% CI −0.6 to 0.9); P=0.77. - Safety: No increase in desaturation events observed with ramelteon; no signal of delayed recovery or prolonged PACU stay.

The trial directly addressed whether preoperative ramelteon (0.1 mg·kg⁻¹) prevents emergence agitation after general anaesthesia in children undergoing tonsillectomy. The incidence of agitation, as well as PAED and Aono’s scores, were indistinguishable between ramelteon and placebo, indicating no preventive effect at this dose. Secondary outcomes, including pain (CHEOPS), preoperative behaviour (PAB), postoperative vomiting, need for rescue medications, recovery time, and PACU discharge rates, also showed no differences, suggesting no ancillary benefits of this dosing regimen. These findings contrast with prior meta-analytic evidence for melatonin, which suggested dose-dependent prevention of emergence agitation (greater effect at 0.5 mg·kg⁻¹). Although ramelteon has higher receptor affinity than melatonin and has shown delirium-preventive effects in adults, the tested paediatric dose (0.1 mg·kg⁻¹) was ineffective for emergence agitation in this high-risk tonsillectomy population. Importantly, ramelteon did not increase respiratory depression or desaturation, supporting its safety in this context. The authors note that higher doses may be necessary to achieve efficacy and recommend further trials to evaluate dose-response while maintaining vigilant respiratory monitoring. Robust blinding and allocation concealment are strengths that reduce bias, but the single-centre, procedure-specific design may limit generalizability.

Ramelteon at 0.1 mg·kg⁻¹ did not prevent emergence agitation after general anaesthesia in paediatric patients undergoing tonsillectomy and did not improve pain, behaviour, postoperative vomiting, recovery time, or PACU stay. The regimen appeared safe without increased desaturation. Future research should investigate higher dosing strategies of ramelteon for preventing emergence agitation in children and assess applicability across different surgical procedures and age strata.

- Single-centre study limited to children undergoing tonsillectomy, a procedure associated with a high baseline risk of emergence agitation; results may not generalize to other surgeries or settings. - Only one dose (0.1 mg·kg⁻¹) of ramelteon was evaluated; higher doses may be necessary for efficacy but were not tested. - Wide age range (18–119 months) may introduce variability in behavioural responses and agitation risk. - Potential ceiling effect due to high agitation incidence in tonsillectomy could have obscured a modest treatment effect. - Low preoperative behaviour scores in the control group may have limited detection of anxiolytic benefits. - Three participants did not receive premedication (protocol deviations), though included in modified ITT analysis.