Purpose: Sleep disturbance is a prevalent problem for cancer survivors and effective behavioral treatments are not widely used for this population. This study evaluated home-based sleep interventions based on cognitive behavioral therapy for insomnia (CBT-I).
Methods: This phase II randomized controlled trial evaluated two manualized interventions over 7 weeks. The intervention group received sleep hygiene information, stimulus control, sleep restriction, and a bedtime imagery audio recording. The control group was similar, but without sleep restriction and used audio recordings of bedtime short stories instead of imagery. Eligibility included adult cancer survivors who had trouble falling asleep or falling back to sleep on 3 of 7 days. Patients with diagnoses of sleep or mental health disorders were excluded. The primary endpoint was change in time to fall asleep or falling back to sleep after awakening, from baseline to week 7. Two-sample T tests evaluated differences between arms for this endpoint.
Results: Ninety-three of 168 planned participants were enrolled from 20 institutions. The study closed early for poor accrual. Baseline time to sleep was 45 min and 52 min for the intervention and control group, respectively. At 7 weeks, both groups improved, the intervention group to 26 min and control group to 30 min, a non-significant difference between groups (p = 0.85). Secondary outcomes improved in both groups with no significant differences between arms.
Conclusions: Improvement in sleep outcomes in both arms was consistent with other CBT-I interventions delivered through alternative approaches to provider-delivered therapy. More research on optimal scalable delivery of CBT-I is needed.
Clinical relevance: This study supports the effectiveness of CBT-I based behavioral interventions for sleep but also the need for better delivery methods to improve uptake and effect size.
Trial registration: ClinicalTrials.gov identifier: NCT00993928.
Keywords: Behavioral intervention; CBT-I; Cancer survivors; Sleep disturbance.