The cost-effectiveness of an online intervention to prevent dementia: results from the Maintain Your Brain (MYB) randomised controlled trial

BACKGROUND: The Maintain Your Brain (MYB) randomised controlled trial (RCT) examined the effect of a multi-domain internet-based dementia prevention program against a control group (information only). OBJECTIVES: A cost-effective analysis (CEA) quantified the differences in costs (direct healthcare and program costs) and effectiveness outcomes between the intervention and control groups from a healthcare sector perspective. DESIGN: An economic evaluation was conducted alongside the MYB RCT over three years. SETTING: Australians aged 55-77 years with at least 2 identified remediable risk factors for cognitive decline/dementia recruited from communities in New South Wales. PARTICIPANTS: There were 3,025 participants in the intervention group and 3,033 in the control group with available linked healthcare data via the Sax Institute's 45 and Up Study out of the 6104 enrolled in the trial (99.2% of total cohort). INTERVENTION: The MYB trial comprised a personalised schedule of online coaching in physical activity, nutrition, cognitive activity, and depression or anxiety management. MEASUREMENTS: The two effectiveness outcomes were global cognition composite (GCC) scores and the Australian National University-Alzheimer's Disease Risk Index -short form (ANU-ADRI-SF) questionnaire scores. Costs included MYB program costs and the direct healthcare costs incurred by the MYB participants. All costs were reported in Australian dollars (AUD$) during the trial period. The time horizon of this analysis was 3 years after randomisation (2018-2021). Incremental cost-effectiveness ratio (ICERs) between the intervention and the control groups were calculated by comparing the average difference in costs to a mean difference in z score for GCC and ANU-ADRI-SF score using the bootstrapped means and 95% Confidence Intervals. RESULTS: The total unadjusted program and healthcare costs over three years were similar between groups (AUD$16,521 per person in the control group and AUD$16,473 in the intervention group). After adjusting for baseline characteristics, the average difference between groups in total cost per person at three years was not statistically different: AUD$467 favouring the control group (95%CI: -$552 - $1585). This was compared to a significant mean difference (improvement) in GCC z score at three years of 0.18 (95%CI: 0.13, 0.23) and -0.57 (95%CI: -0.95, -0.24) point difference in ANU-ADRI-SF for the intervention versus control. The base case ICERs were AUD$2,568 per 1 standard deviation in z score and $823 per reduction of 1 ANU-ADRI-SF point. With 1000 bootstrapped replications, the scatterplots of ICER ellipses suggest that the MYB intervention was more effective than the control group and with no significant difference in overall healthcare costs. CONCLUSION: The MYB trial showed cost-effectiveness for preventing cognitive decline and reducing dementia risk. Longer-term follow-up and dissemination to other cohorts is needed to confirm the impact on preventing future cases of dementia and relevance to other socio-economic and cultural/ethnic groups than those enrolled in the original trial.