Abstract
Central nervous system (CNS)-active medications pose serious health risks for older adults with dementia but are nonetheless commonly used. Few deprescribing interventions have focused on people with dementia. We conducted a one-arm pilot study in six primary care practices of an integrated healthcare system between February and August 2023. The deprescribing intervention consisted of patient/care partner education and self-management materials and provider decision support. Participants were aged 60+ with diagnosed dementia and prescribed at least one CNS-active medication for three or more months of the six-month period prior to study start. We assessed feasibility and acceptability of the intervention and feasibility of ascertaining medication discontinuation and medically treated falls. The intervention was delivered to all (N = 114) eligible participants; their mean age was 80 ± 9 years; 72% were female and 13% non-White. Intervention acceptability, assessed by Weiner's Acceptability of Intervention measure, was rated 3.5/5 (range 1-5; higher scores indicate higher acceptability). Among baseline antipsychotic users (N = 89), 39 (43.8%) had discontinued at follow-up. Among baseline tricyclic antidepressant users (N = 11), 6 (54.5%) had discontinued at follow-up. Among baseline skeletal muscle relaxant users (N = 3), 2 (66.7%) had discontinued at follow-up. Among baseline benzodiazepine users (N = 3), 1 (33.3%) had discontinued at follow-up. Among baseline opioid users (N = 13), 1 (7.7%) had discontinued at follow-up. Medically treated falls occurred among 22% at baseline vs. 21% at follow-up. The intervention is feasible and acceptable and may achieve meaningful reduction in CNS-active medication prescriptions. Findings support a controlled trial with sufficient power to assess effects on relevant clinical outcomes.