INTRODUCTION: First-line treatment for behavioral and psychiatric symptoms of dementia is non-pharmacological. Still, psychotropic medication is widely used, despite its limited effect and harmful side-effects. More than half of all nursing home residents with dementia receive antidepressants, even though deprescribing is safe and feasible. Interventions to promote deprescribing of antidepressants in nursing homes are few and complex. To optimize the deprescribing process through an intervention, transparency for the development of the intervention is needed. We aim to describe the steps in the development and tailoring of an intervention targeting GPs, nursing home staff, and relatives to enhance collaboration on reducing the use of antidepressants in institutionalized older persons with dementia in Denmark.

METHOD: A step-wise process guided by the core elements in the Medical Research Council constituted the tailoring process. Five steps were included; 1) a literature search, 2) interviews with stakeholders, 3) drafting the intervention prototype, 4) professionals' assessment of the intervention, and 5) refinement of the intervention. The steps were conducted from June 2020 to June 2022.

RESULTS: Based on the literature search, interviews with stakeholders, and professionals' assessment of the intervention, four main themes were identified; 1) focusing on antidepressants, 2) importance of professional qualifications, 3) collaboration and communication, and 4) patient and relative involvement. They guided intervention development and refinement of the final intervention, which included 1) a case-based training course and 2) a dialog tool including a symptom assessment scale to be used in a structured consultation at the nursing home.

CONCLUSION: This study presents a detailed account of the tailoring process for a complex intervention to optimize deprescribing of antidepressants for older persons with dementia at nursing homes. By presenting a thorough development process, we expect to achieve increased adherence to the intervention which is currently being tested in an ongoing cluster randomized controlled trial. The transparency of the process will also increase the future development of other similar complex interventions.