0421 Synchronous Virtual Cognitive Behavioral Therapy for Insomnia: Preliminary Results of Sleep Pattern Changes among Cancer Survivors

Abstract Introduction For cancer survivors, insomnia is a frequent and distressing concern. Cognitive behavioral therapy for insomnia (CBT-I) is an effective treatment, yet in-person care can be difficult for cancer survivors to access for a variety of socioeconomic, medical and logistical reasons. In a pilot randomized clinical trial, four, weekly sessions of virtually-delivered, synchronous CBT-I were feasible, acceptable, and seemingly efficacious for reducing insomnia among cancer survivors (vs. an enhanced usual care control). The aim of this secondary analysis was to characterize daily patterns in sleep diary metrics during the intervention phase. Methods Adults who had completed treatment for any non-metastatic cancer and had chronic insomnia (DSM-5 criteria and Insomnia Severity Index score <=15) and were randomized to receive CBT-I completed daily assessments of sleep quality (SQ), total sleep time (TST), time in bed (TIB), sleep efficiency (SE), sleep onset latency (SOL), and waking after sleep onset (WASO) for 21 days (i.e., from Session 1-4). Linear mixed models were used to compute trajectories of change for each sleep metric and -2 log likelihood, AIC, and BIC fit indices. Results Among 20 cancer survivors who received CBT-I (mean age 51.2 ± 17.1 years, 90% female), model fit indices indicated that across sleep metrics, linear growth models provided a better fit to the data than either quadratic or cubic growth models. Significant linear improvement was observed for SQ (B = 0.02, t = 3.60, p < 0.001), SE (B = 0.16, t = 2.01, p = 0.046), and SOL (B = -0.41, t = -2.47, p = 0.01), but not TIB, TST, or WASO (ts < 1.12, ps > 0.26). Conclusion Findings elucidate how sleep may be expected to change over the course CBT-I, in particular for the growing population of cancer survivors seeking virtual care for insomnia. The strengths, limitations, and potential modifications to optimize the intervention across sleep metrics will be discussed. Support (if any) This work was conducted with support from the American Cancer Society (Massachusetts General Hospital Institutional Research Grant, PI: Hall).