Is a definitive trial of Take-Home Naloxone in emergency settings indicated? Results of a cluster randomised feasibility study

Abstract Background Opioids kill more people than any other class of drug. Naloxone is an opioid antagonist which can be distributed in kits for peer administration. We aimed to determine feasibility of undertaking a definitive randomised controlled trial (RCT) of Take-home Naloxone (THN) in emergency settings. Methods Using individual-level-routine health records (2015-21) we tested feasibility of developing a discriminant function to identify people at high-risk of fatal opioid poisoning for outcome comparisons. We undertook a clustered RCT on paired UK Emergency Department (ED) and ambulance service sites. At intervention sites, we recruited practitioners to administer THN to patients presenting with opioid overdose or related condition during ta 1year recruitment period, 2019 – 21. We assessed feasibility of intervention and trial methods against predetermined progression criteria. Results Within routine health records on the population of Wales (~3,200,000), we identified 1,105 adult deaths from opioid poisoning, of whom 307 (27.8%) had no ED or drugs service contacts in the year before death. At a predicted probability threshold of 0.0003, a discriminant function based on demographics and recent healthcare contacts identified 809 opioid related deaths within 1 year (sensitivity 74.7%) in 989,151 people, missing 274 cases. Lowering the threshold to 0.0002 increased sensitivity to 86.1% but included a further 608,191 non-cases; raising it to 0.0004 reduced sensitivity to 65.4% and inclusion of non-cases to 646,750. At two intervention sites, randomly selected from 4: 299/687 (43.5%) clinical staff were trained; 60/277 eligible patients (21.7%) were supplied with a THN kit and no adverse events were reported. Conclusion With a low incidence of opioid-related death and significant proportion with no contact with ED or drug services in the year before death, the numbers needed to reach a reasonable sensitivity was very high. This study did not meet progression criteria, a fully powered trial is not planned. Trial Registration ISRCTN13232859 (Registered 16/02/2018)