Deprescribing benzodiazepines and Z-drugs in community-dwelling adults: a scoping review

André S. Pollmann, Andrea L. Murphy, Joel C. Bergman, David M. Gardner

Long-term sedative use is prevalent and associated with significant morbidity, including adverse events such as falls, cognitive impairment, and sedation. The development of dependence can pose significant challenges when discontinuation is attempted as withdrawal symptoms often develop. We conducted a scoping review to map and characterize the literature and determine opportunities for future research regarding deprescribing strategies for long-term benzodiazepine and Z-drug (zopiclone, zolpidem, and zaleplon) use in community-dwelling adults. We searched PubMed, Cochrane Central Register of Controlled Trials, EMBASE, PsycINFO, CINAHL, TRIP, and JBI Ovid databases and conducted a grey literature search. Articles discussing methods for deprescribing benzodiazepines or Z-drugs in community-dwelling adults were selected. Following removal of duplicates, 2797 articles were reviewed for eligibility. Of these, 367 were retrieved for full-text assessment and 139 were subsequently included for review. Seventy-four (53 %) articles were original research, predominantly randomized controlled trials (n = 52 [37 %]), whereas 58 (42 %) were narrative reviews and seven (5 %) were guidelines. Amongst original studies, pharmacologic strategies were the most commonly studied intervention (n = 42 [57 %]). Additional deprescribing strategies included psychological therapies (n = 10 [14 %]), mixed interventions (n = 12 [16 %]), and others (n = 10 [14 %]). Behaviour change interventions were commonly combined and included enablement (n = 56 [76 %]), education (n = 36 [47 %]), and training (n = 29 [39 %]). Gradual dose reduction was frequently a component of studies, reviews, and guidelines, but methods varied widely. Approaches proposed for deprescribing benzodiazepines and Z-drugs are numerous and heterogeneous. Current research in this area using methods such as randomized trials and meta-analyses may too narrowly encompass potential strategies available to target this phenomenon. Realist synthesis methods would be well suited to understand the mechanisms by which deprescribing interventions work and why they fail.