An investigation into the administration of haloperidol, levomepromazine and midazolam to inpatients dying of COVID-19 at a tertiary hospital during the coronavirus pandemic in the UK

Background: In April 2020, UK national and local guidelines for symptom control at the end of life were revised to indicate that higher doses of anti-agitation medication (AAM) might be required for patients dying due to COVID-19 . Aims: To investigate AAM ( haloperidol, levomepromazine or midazolam) doses administered to inpatients dying of COVID-19 at a tertiary hospital during the first wave of the pandemic. Methods: Electronic prescribing records of hospital inpatients who died of COVID-19 between 15th March and 15th June 2020 were examined (excluding patients dying on intensive care unit). Doses of AAM administered in the last 48 hours of life were recorded anonymously from electronic prescribing records, including administration route, and analysed by descriptive statistics. The School of Pharmacy Ethics committee granted ethical approval. Results: Included inpatients (n=258), were predominantly male (55%) and White British (70%). In total, 132 inpatients received at least one AAM via any administration route. The majority of inpatients (n=116) received at least one dose of 'as required' AAM as a subcutaneous injection. 92% of those inpatients received midazolam (median dose 5mg), 24% received levomepromazine (median dose 6.25mg) and 3% received haloperidol (median dose 2mg). For inpatients receiving at least one AAM via continuous subcutaneous infusion (n=55), 89% received midazolam (median dose 10mg/24 hours), 16% received levomepromazine (median dose 12.5mg/24 hours), and 5% received haloperidol (median dose 2.5mg/24 hours). Conclusion: Our results indicate that doses of AAMs administered to our inpatients dying of COVID-19 were more in line with 'usual practice' and not as high as anticipated in view of the revised guidance. Our next step is to interpret these results in the context of the dying patients' actual symptom burden.