Effects of a Triage Checklist to Optimize Insomnia Treatment Outcomes and Reduce Hypnotic Use: The RESTING Study.

Abstract

Evaluate a triaged stepped-care strategy among adults 50 and older with insomnia disorder. Participants (N=245) were classified at baseline by a Triage-Checklist. Those projected to do better if they start treatment with therapist versus digitally delivered CBT-I (tCBT-I versus dCBT-I) constituted the YES stratum (n=137); the rest constituted the NO stratum (n=108). Participants were randomized within stratum to a strategy that utilized only dCBT-I (ONLN) or to a strategy that prospectively allocated the first step of care to dCBT-I or tCBT-I based on the Triage-Checklist and switched dCBT-I non-responders at 2-months to tCBT-I (STEP). Co-primary outcomes were the insomnia severity index (ISI) and the average nightly amount of prescription hypnotic medications used (MEDS), assessed at 2,4,6,9, and 12 months post-randomization. Mixed effects models revealed that, compared to ONLN, participants in STEP had greater reductions in ISI (p=0.001; η2=0.01) and MEDS (p=0.019, η2=0.01). Within the YES stratum, compared to ONLN, those in STEP had greater reductions in ISI (p=0.0001, η2=0.023) and MEDS (p=0.018, η2=0.01). Within the ONLN arm, compared to the YES stratum, those in the NO stratum had greater reductions in ISI (p=0.015, η2=0.01) but not in MEDS. Results did not change with treatment-dose covariate adjustment. Triaged-stepped care can help guide allocation of limited CBT-I treatment resources to promote effective and safe treatment of chronic insomnia among middle age and older adults. Further refinement of the Triage-Checklist and optimization of the timing and switching criteria may improve the balance between effectiveness and use of resources.