An online behavioural sleep intervention (COSI) added to standard care for children with epilepsy showed significant superiority of standard care alone on the primary outcome, though children in the COSI group showed a mean 16.5-minute reduction in sleep onset latency by actigraphy and parents increased their knowledge.
Key Findings
Results
Standard care alone was significantly superior to standard care plus COSI on the primary outcome measure of children's sleep habits at three months.
The primary outcome was the Children's Sleep Habits Questionnaire (CSHQ) at three months.
The adjusted mean CSHQ difference between arms was 3.00 (95% CI 0.06–5.93; p = 0.05), indicating significant superiority of SC.
85 children were enrolled (42 SC; 43 SC + COSI) across 26 UK outpatient clinics.
Children were aged 4–12 years with epilepsy and sleep problems.
Analyses were conducted on an intention-to-treat basis.
Results
Children in the SC + COSI group showed a mean 16.5-minute reduction in sleep onset latency as measured by actigraphy.
Sleep onset latency reduction of 16.5 minutes was an objective finding measured via actigraphy.
This improvement was observed despite the primary CSHQ outcome favouring standard care alone.
The authors noted this suggests 'the underlying behaviour change techniques hold value.'
Results
Only 53% of families assigned to the COSI intervention accessed the core intervention materials.
Only 23 (53%) families in the SC + COSI group accessed the core intervention materials.
This low engagement rate may have contributed to the lack of superiority on the primary outcome.
The trial was multicentre, parallel-group, and unblinded.
Results
Parents of children in the COSI group increased their knowledge about sleep.
Enhanced parental knowledge was reported as a positive finding for the SC + COSI arm.
This was noted alongside the objective improvement in sleep onset latency as evidence that 'the underlying behaviour change techniques hold value.'
Results
The SC + COSI intervention was not cost-effective at the standard UK threshold of £30,000 per QALY.
The incremental mean cost of SC + COSI was £1,232 (95% credibility interval £535–£3,455).
The mean incremental Quality Adjusted Life Year (QALY) was 0.00 (95% CI −0.03 to 0.04).
This yielded an incremental cost-effectiveness ratio of £433,167 per QALY gained.
There was only a 0.04 probability of the intervention being cost-effective at the £30,000/QALY threshold.
Cost-effectiveness was estimated at six months.
Methods
The trial was a multicentre, parallel-group, unblinded randomised controlled trial recruiting children with epilepsy and sleep problems from 26 UK outpatient clinics.
Participants were randomly assigned (1:1) via a computer-generated minimisation algorithm.
Children were aged 4–12 years with epilepsy and sleep problems.
Recruitment occurred through 26 UK outpatient clinics.
The intervention evaluated was an online sleep intervention (COSI) for parents.
What This Means
This research evaluated an online behavioural sleep program called COSI, designed to help parents of children with epilepsy manage their children's sleep problems. The study randomly assigned 85 children (aged 4–12 years) to either receive standard medical care alone or standard care plus access to the online COSI program. The main measure of success was a questionnaire about children's sleep habits completed at three months. Contrary to what the researchers hoped to find, standard care alone actually performed significantly better than the combination of standard care plus COSI on this questionnaire.
However, there were some encouraging signs within the results. Children whose families received COSI fell asleep about 16.5 minutes faster on average, as measured by wrist-worn activity monitors (actigraphy), and their parents gained more knowledge about sleep. A major challenge was that only about half (53%) of the families assigned to COSI actually accessed the core program materials, which likely limited how much benefit the intervention could provide overall.
From a cost perspective, the COSI program was not considered good value for money by UK health standards — the estimated cost per quality-adjusted life year (a standard health economics measure) was over £433,000, far above the usual £30,000 threshold used to judge cost-effectiveness. This research suggests that while the behaviour-change techniques embedded in the COSI program may have some genuine value — particularly for improving how quickly children fall asleep and for educating parents — the online delivery format as tested here faces significant challenges with family engagement and overall cost-effectiveness for this particular population.
Gringras P, Anilkumar A, Bray L, Carter B, Coffey T, Cook G, et al.. (2026). Randomised controlled trial of online behavioural sleep intervention for children with epilepsy.. Scientific reports. https://doi.org/10.1038/s41598-025-27206-3