Cost-utility analysis of the DREAMS START intervention for people living with dementia and their carers: a within-trial economic evaluation.

BACKGROUND: People living at home with dementia frequently have disturbed sleep. The multicomponent, non-pharmacological intervention DREAMS START has shown to be effective at improving sleep in this population. We aimed to conduct a cost-utility analysis of DREAMS START compared with treatment as usual (TAU). METHODS: This economic evaluation within a single-masked, phase 3, parallel-arm, superiority randomised controlled trial involved dyads of people with dementia and sleep disturbance and their family carer. Participants were recruited from the National Health Service and the Join Dementia Research service in England. Dyads were randomly assigned (1:1) to receive the DREAMS START intervention (plus TAU) or TAU. Randomisation was blocked, with stratification by site, and a web-based system was used for allocation. Researchers collecting outcome data were masked to allocation group. The primary outcome was sleep disturbance measured by the Sleep Disorders Inventory (SDI) at 8 months. At baseline, 4 months, and 8 months, family carers completed the 5-level EuroQoL 5 dimensions (EQ-5D-5L) proxy, the Dementia Quality of Life Instrument (DEMQOL)-Proxy, and EQ-5D-5L questionnaires, and resource use for the patient and family carer was measured. We calculated the probability that the DREAMS START intervention is cost-effective from a health and personal social services perspective and from a wider societal perspective for a range of decision thresholds per quality-adjusted life-year (QALY) gained using the EQ-5D-5L scores to calculate QALYs and imputing missing data, reported with a cost-effectiveness acceptability curve. This trial was registered with ISRCTN, 13072268, and is complete. FINDINGS: From Feb 24, 2021, to March 5, 2023, we randomly assigned 377 dyads: 188 to the intervention group and 189 to the TAU group. The mean age of participants with dementia was 79⋅4 years (SD 9⋅0), 206 (55%) of whom were women and 171 (45%) were men. As previously reported, the mean SDI score at 8 months was lower in the intervention group than in the TAU group (adjusted difference in means -4·70 [95% CI -7·65 to -1·74], p=0·002). The mean incremental difference in health and personal social services costs was £59 less per dyad (95% CI -5168 to 5050) and, when incorporating wider societal costs, was £116 less per dyad (-5769 to 5536) for the intervention group than the TAU group, although these figures were non-significant. The mean incremental difference in QALYs per person with dementia was 0·016 more (95% CI 0·000 to 0·033) for the intervention group than for the TAU group, indicating no significant difference in quality of life. At a £20 000 per QALY gained decision threshold, there was a 78% probability that DREAMS START is cost-effective, compared with TAU. INTERPRETATION: DREAMS START is likely to be cost-effective. Given its clinical effectiveness, we recommend that this intervention forms part of routine care for people with dementia and disturbed sleep. FUNDING: National Institute for Health and Care Research Health Technology Assessment.