Melatonin versus midazolam in the premedication of anxious children attending for elective surgery under general anaesthesia: the MAGIC non-inferiority RCT.

Background: Anxiety in children prior to general anaesthesia is common, with up to half displaying distress. Anxiety and distress may lead to unsuccessful anaesthesia, together with greater postoperative pain, agitation and behavioural changes after surgery including sleep disturbances. Midazolam is the current standard premedication; however, it has adverse effects such as the potential for respiratory suppression and unpredictable effects which may result in agitation rather than anxiolysis. Melatonin is an alternative preoperative anxiolytic; however, previous trials have delivered conflicting results. The aim of this non-inferiority trial was to evaluate the effectiveness of melatonin compared to midazolam in reducing anxiety in children undergoing general anaesthesia. Methods: We undertook a randomised-controlled, parallel-group, double-blind, non-inferiority trial in 20 United Kingdom National Health Service trusts, with an embedded qualitative study and health economic evaluation. Anxious children having day case elective surgery under general anaesthesia were randomly assigned to either control (standard of care) group: midazolam; or intervention group: melatonin. The primary outcome was preoperative distress (non-inferiority hypothesis) as assessed by modified Yale Preoperative Anxiety Scale Short Form. Secondary outcomes included safety and efficacy objectives. Analyses were by intention to treat, with an additional per-protocol analysis. The sample size of the trial was 624 children. Results: The trial was stopped early due to recruitment futility. Between 30 July 2019 and 9 November 2022, 110 children were recruited; 55 allocated to midazolam and 55 allocated to melatonin. Pre-planned analyses showed an adjusted mean difference of 13.1 (95% confidence interval 3.7 to 22.4) for the intention-to-treat population and 12.9 (95% confidence interval 3.1 to 22.6) for the per-protocol population, in favour of midazolam. In both analyses, the upper limit of the 95% confidence interval exceeds the predefined margin of 4.3; therefore, melatonin is not non-inferior to midazolam. The lower limit of the 95% confidence intervals excludes zero and thus melatonin is inferior to midazolam; the difference found is considered to be clinically meaningful. Adverse events in the midazolam arm (26%) were slightly higher than melatonin (18%); there were no serious adverse events in either arm. Challenges to recruitment included study-related factors (eligibility criteria and trial design), participant factors (caregiver stress on the day of treatment) and practitioner factors (valuing predictability). In terms of acceptability, preferences of the anaesthetist, patient and caregiver factors and medication side effects profile were influential and suggest the choice of preoperative anxiolytic is more complex than previously described. On average, costs over the 14 days post surgery were lower for those who received melatonin (-£46.20, 95% confidence interval -£166.14 to £66.74) with a mean incremental difference in procedure success of -0.02 (95% confidence interval -0.08 to 0.004), although there was uncertainty around the results. Conclusion: In children with preoperative anxiety, midazolam is more effective than melatonin at reducing preoperative anxiety prior to general anaesthesia, although the early termination of the trial increases the likelihood of bias. Limitations: The trial was prematurely terminated due to recruitment futility. Despite this, a clinically meaningful and statistically significant finding was observed about the primary outcome. Future work: There remains a need to develop or repurpose another drug with a more favourable side effects profile to midazolam. Funding: This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number 16/80/08.