While almost half of older adults admitted to hospital are prescribed potentially inappropriate medicines, less than 1% have a medicine proactively deprescribed during admission in the UK. The CompreHensive geriAtRician-led MEdication Review (CHARMER) intervention is designed to address geriatricians’ and pharmacists’ barriers and enablers to deprescribing. The CHARMER definitive trial will evaluate effectiveness, cost-effectiveness and safety.

A stepped-wedge cluster randomised controlled trial will be conducted in 20 hospitals in England, with four hospitals in reserve. All hospitals will collect baseline data. Every 3 months, five hospitals will be randomised to receive the intervention. The intervention, implemented by a local project manager, comprises a hospital action plan to set deprescribing as an organisational goal; workshops for pharmacists and geriatricians to change beliefs about deprescribing; weekly briefings between geriatricians and pharmacists to discuss opportunities for deprescribing; benchmarking reports to compare deprescribing performance across participating hospitals. With an average of 200 patients admitted and discharged during each step, the study will have 89.5% power at 5% significance level and intra-class correlation coefficient of 0.05 to detect a 3% difference in 90-day re-admission rate from 16.7% versus 13.7%. Anonymised routinely collected data, including readmissions, will be obtained for all patients admitted during the study period. Enhanced data collection periods of 1 month during control and intervention periods will be used to recruit patients and data for secondary outcomes and process evaluation.

A stepped-wedge design enabled a smaller number of hospitals and patients to be included than a traditional cluster-randomised design. The complexity of intervention implementation necessitated a project manager in addition to the principal investigator responsible for trial conduct. Using routinely collected data for the primary outcome measure should ensure that the trial has sufficient power on completion. Planned enhanced data collection for short periods of time improves trial efficiency.

ISRCTN13248281.