Study in Outpatient Medicine using Nudges to improve Sleep: The SOMNUS Trial - ABSTRACT Sedative-hypnotic z-drugs are dependence-forming substances associated with falls, fractures, high-risk parasomnias (sleepwalking, sleep-driving), and cognitive impairment. However, these drugs are popular and are usually prescribed long-term for insomnia, despite guideline recommendations to limit their use and to favor cognitive-behavior therapy for insomnia. Behavioral economic nudges delivered in the electronic health record are a promising way to encourage the use of cognitive-behavioral bibliotherapy and phone apps for insomnia and to reduce high-risk Z-drug prescribing. Nudges are insights from behavioral science that fit into existing clinical workflows, intended to preserve choice but influence decision-making toward desired behavior. We propose the Study in Outpatient Medicine using NUdges to improve Sleep (SOMNUS) in response to NHLBI's Notice of Special Interest (NOT-HL-21-010). The SOMNUS trial will simulate the long-term outcomes of our nudging strategy using microsimulation from our Roybal Center for Behavioral Interventions in Aging's Future Americans Model (Aim 1), align workflow and nudge performance in the electronic health record to develop the best choice architecture (Aim 2), and conduct a randomized trial of minimally disruptive changes to the electronic health record to encourage the use of evidence-based therapies for insomnia in everyday clinical practice at 60 clinics in the U.S. (Aim 3). The trial will employ the following two nudge concepts: 1) Social Accountability Nudges that ask a clinician seeking to prescribe a Z-drug to either justify, in a brief free-text response, the reason for a guideline-discordant action or to cancel the order, and to precommit to a brief deprescribing plan whenever prescribing a Z-drug; 2) setting default days supply, pre-populated in the medication order, to a small number of pills. In an 18-month trial involving 60 clinics (449 clinicians), clinic-clusters will be randomized to receive a) social accountability nudges, b) defaults, c) both social accountability and defaults, or d) no intervention (education control). The primary outcome measure is the change in Z-drug days supply ordered per patient. Secondary outcomes include probability of starting a new patient on a Z-drug, probability of discontinuation among long-term users, and a physician perceptions survey. To aid in deprescribing Z-drugs, we will also provide bibliotherapy, access to insomnia, referrals for screening for related problems and visit assessments. Sample size calculations indicate an 88.5% (95% CI: 83.3, 92.6) power to detect a 6.7% decrease in pills ordered. This proposal will integrate behavioral economic approaches into implementation research to encourage evidence-based decisions involving the treatment of insomnia. It will improve the safety and efficacy of everyday clinical practice for people visiting their physician with sleep problems. It will also provide estimates of the benefits of treatment to future Americans.
