0399 Masked Tapering and Augmented CBTI Improves Benzodiazepine Receptor Agonist Discontinuation in a Randomized Trial

Abstract Introduction Benzodiazepines and benzodiazepine receptor agonists (BZRAs) are not recommended as first-line therapy for insomnia in older adults due to adverse events (e.g., falls). Although guidelines recommend cognitive behavioral therapy for insomnia (CBTI) and BZRA discontinuation, discontinuing BZRAs is challenging. To improve discontinuation success, we developed and tested a novel intervention (grounded in the science of known placebo effects of BZRAs), masked taper plus cognitive behavioral therapy-augmented program (MTcap), which masks the daily BZRA dose during tapering and adds novel cognitive and behavioral exercises targeting placebo effects to CBTI. We hypothesized that MTcap would increase BZRA discontinuation 1 week (PTX) and 6 months post-treatment (6M) compared to standard CBTI plus supervised (unmasked) gradual taper (SGT). Methods In a multi-site randomized trial, adults (>= 55 years) who use lorazepam, alprazolam, clonazepam, temazepam, and/or zolpidem at doses <= 8 mg diazepam-equivalent 2+ nights/week for >= 3 months for insomnia were randomized to MTcap versus SGT. Blinded research staff conducted PTX and 6M assessments including one-week medication logs and Insomnia Severity Index (ISI). Effects of MTcap versus SGT on BZRA discontinuation (primary outcome) were modeled with logistic regression. Using two-level mixed-effects models, we predicted additional outcomes (frequency [#days/week taken], dose, ISI) as a function of treatment group, site, treatment-site interaction, and time. Results 188 participants (mean age 69.2 years, 34.6% female, mean frequency 5.9 days/week BZRA use, mean diazepam-equivalent 3.9 mg, mean ISI 14.0) were randomized (MTcap n=92, SGT n=96), with PTX and 6M follow-up rates of 94.7% and 93.6%, respectively. BZRA discontinuation for MTcap was superior to SGT at both PTX (MTcap=0.884, SGT=0.674; odds ratio [OR] 3.683, 95% CI 1.670, 8.122, p=.001) and 6M (MTcap=0.734, SGT=0.586; OR 1.955, 95% CI 1.033, 3.700, p=.039). Change in BZRA frequency was lower in MTcap than SGT at PTX (-1.32, 95% CI -2.07, -0.57, p<.001), but not 6M. ISI did not differ between these two active treatment groups at PTX or 6M. Conclusion Adding a masked taper and novel cognitive and behavioral exercises targeting placebo effects to traditional unmasked tapering plus CBTI results in markedly improved long-term BZRA discontinuation, with similar improvement in symptoms of insomnia. Support (if any) VAIIR 17-234, NIAR01AG057929, UL1TR001881