Improving sleep after stroke: A randomised controlled trial of digital cognitive behavioural therapy for insomnia

Abstract

Abstract Stroke is frequently accompanied by long-term sleep disruption. We therefore aimed to assess the efficacy of digital cognitive behavioural therapy (dCBT) for insomnia to improve sleep after stroke. A parallel group randomised controlled trial was conducted remotely in participant’s homes/online. Randomisation was online with minimisation of between-group differences in age and baseline Sleep Condition Indicator (SCI-8) score. In total, 86 community dwelling stroke survivors consented, of whom 84 completed baseline assessments (39 female, mean 5.5 years post-stroke, mean 59 years of age) and were randomised to dCBT or control (sleep hygiene information). Follow-up was completed at post-intervention (mean 75 days after baseline) and 8-weeks later. The primary outcome was self-reported insomnia symptoms, as per the SCI-8 (range 0-32, lower numbers indicate more severe insomnia, reliable change 7-points) at post-intervention. There were significant improvements in SCI-8 for dCBT compared to control (intention-to-treat, dCBT n=48, control n=36, 5 imputed datasets, effect of group p≤0.02, η p 2 =0.07–0.12, pooled mean difference=-3.35). Additionally, secondary outcomes showed shorter self-reported sleep onset latencies and better mood for the dCBT group, but no significant differences for self-efficacy, quality of life, or actigraphy-derived sleep parameters. Cost-effectiveness analysis found that dCBT dominates over control (non-significant cost savings and higher quality adjusted life years). No related serious adverse events were reported to the researchers. Overall, dCBT for insomnia effectively improves sleep after stroke. Future research is needed to assess earlier stages post-stroke, with a longer follow-up period to determine whether it should be included as part of routine post-stroke care. Clinicaltrials.gov NCT04272892

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 33 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.