The MAGIC trial: a pragmatic, multicentre, parallel, noninferiority, randomised trial of melatonin versus midazolam in the premedication of anxious children attending for elective surgery under general anaesthesia
Robert Bolt, Marie Hyslop, Esther Herbert, Diana Papaioannou, Nikki Totton, Matthew Wilson, Jan Clarkson, Christopher Evans, Nicholas Ireland, Jennifer Kettle, Zoe Marshman, Amy Norrington, R. Paton, Christopher R. Vernazza, Chris Deery, Sondos Albadri, Laura Marie Armstrong, Simon Atkins, Margaret Babb, Claire Biercamp, Katie Biggs, Mike Bradburn, Jaimie Buckley, Julie Child-Cavill, Sean Cope, Simon Crawley, Munyaradzi Dimairo, Enass Duro, Ayman Eissa, Laura Flight, Jacqui Gath, Gil Gavel, Tim Geary, Fiona Gilchrist, Padma Gopal, Jamie Hall, Kate Hutchence, Puran Khandelwal, Pranav Kukreja, Ian Leeuwenberg, James Limb, Amanda Loban, Katie Mellor, Nuria Masip, Anthony Moores, Vimmi Oshan, Edward Pickles, Jaydip Ray, Helen Rodd, Sian Rolfe, Elena Sheldon, R. J. Simmonds, Rachel Smith, Ashok Sundar, Anna Thomason, Simon Waterhouse, Graham Wilson, J.R. Yates, Tracey Young
Abstract
BackgroundChild anxiety before general anaesthesia and surgery is common. Midazolam is a commonly used premedication to address this. Melatonin is an alternative anxiolytic, however trials evaluating its efficacy in children have delivered conflicting results.MethodsThis multicentre, double-blind randomised trial was performed in 20 UK NHS Trusts. A sample size of 624 was required to declare noninferiority of melatonin. Anxious children, awaiting day case elective surgery under general anaesthesia, were randomly assigned 1:1 to midazolam or melatonin premedication (0.5 mg kg−1, maximum 20 mg) 30 min before transfer to the operating room. The primary outcome was the modified Yale Preoperative Anxiety Scale-Short Form (mYPAS-SF). Secondary outcomes included safety. Results are presented as n (%) and adjusted mean differences with 95% confidence intervals.ResultsThe trial was stopped prematurely (n=110; 55 per group) because of recruitment futility. Participants had a median age of 7 (6–10) yr, and 57 (52%) were female. Intention-to-treat and per-protocol modified Yale Preoperative Anxiety Scale-Short Form analyses showed adjusted mean differences of 13.1 (3.7–22.4) and 12.9 (3.1–22.6), respectively, in favour of midazolam. The upper 95% confidence interval limits exceeded the predefined margin of 4.3 in both cases, whereas the lower 95% confidence interval excluded zero, indicating that melatonin was inferior to midazolam, with a difference considered to be clinically relevant. No serious adverse events were seen in either arm.ConclusionMelatonin was less effective than midazolam at reducing preoperative anxiety in children, although the early termination of the trial increases the likelihood of bias.Clinical trial registrationISRCTN registry: ISRCTN18296119.