Feasibility and effectiveness of deprescribing benzodiazepines and Z-drugs: systematic review and meta-analysis.
adults in England
benzodiazepine or Z-drug (BZRA) deprescribing
gradual tapering alone
successful discontinuation in the short and long term, and withdrawal symptoms
Abstract
A total of 2.4 million adults in England were dispensed a benzodiazepine or Z-drug (BZRA) in 2017/18, and more than 250 000 patients in the UK take BZRAs beyond the recommended duration. Deprescribing is a clinician-guided process of withdrawing inappropriate drugs. This review aimed to evaluate the evidence base supporting the feasibility and clinical effectiveness of all forms of deprescribing initiatives used to discontinue long-term (≥ 4 weeks) BZRAs. Systematic review of randomized controlled trials evaluating BZRAs deprescribing among adults in community, primary or outpatient settings. MEDLINE, Embase and PsycINFO were searched from inception to February 2021. Primary outcomes were successful discontinuation in the short (< 4 weeks) or long term (≥ 4 weeks) and the occurrence of withdrawal symptoms, behavioural or psychological symptoms. Studies were categorized as pharmacological or non-pharmacological supported interventions. Study quality was assessed using the Cochrane risk-of-bias tool. Where appropriate, risk ratios (RRs), mean differences and 95% confidence intervals (CIs) were calculated, and Mantel-Haenszel methods using the random-effect meta-analysis was undertaken to calculate summary effect estimates. Ten studies were included (n = 1431 participants). Heterogeneity in study design and effect was observed. Benzodiazepines were successfully deprescribed when gradually tapered with non-pharmacological support compared with gradual tapering alone in the short term (n = 124; RR = 2.02; 95% CI = 1.41, 2.89) and long term (n = 123; RR = 2.45; 95% CI = 1.56, 3.85). Benzodiazepine deprescribing was more successful when supported by non-pharmacological methods versus routine care (n = 189; RR = 3.26; 95% CI = 2.36, 4.51). Quality of evidence reporting effectiveness was very low to low. It may be feasible to deprescribe benzodiazepines depending on the process and support mechanisms employed.