Implementation of a hospital deprescribing behaviour change intervention, the CompreHensive geriAtRician-led MEdication Review (CHARMER) trial: a process evaluation protocol

Introduction Proactive deprescribing is the process of stopping a medicine and comprises four steps: (1) identify a patient for potential stop of a medicine, (2) evaluate a patient for potential stop of a medicine, (3) stop a medicine and (4) monitor after stopping.

The CHARMER (CompreHensive geriAtRician-led MEdication Review) trial is a stepped-wedge design to evaluate the effectiveness and cost-effectiveness of a behaviour change intervention to increase proactive deprescribing in hospitals. The CHARMER intervention comprises a deprescribing action plan, deprescribing briefings, videos of successful deprescribing consultations, deprescribing case studies workshop and a deprescribing performance dashboard. The process evaluation will explore trial processes, CHARMER intervention implementation, CHARMER behavioural mechanisms of action and contextual factors influencing these aspects.

Methods and analysis The convergent parallel design process evaluation will follow the UK Medical Research Council guidance. We will interview: staff involved in CHARMER implementation, geriatricians and pharmacists who receive the intervention and research delivery staff involved in patient/carer recruitment and data collection. We will also interview patients/carers and primary care practitioners. Interviews will be supplemented with recordings of implementation activities and completed implementation manuals. Questionnaires will capture the extent to which the four proactive deprescribing steps are enacted by intervention recipients, measure the behavioural mechanisms by which the CHARMER intervention operates and capture the patient experience of proactive deprescribing. Qualitative data will be analysed thematically and then mapped to Normalisation Process Theory to explore implementation and the Theoretical Domains Framework to explore behaviour change. Most quantitative data will be analysed descriptively; however, changes in staff questionnaire responses preintervention and postintervention will be analysed using a Mann-Whitney test. We will triangulate qualitative and quantitative findings to explain intervention effects.