Within trial cost-utility analysis of disease management program for patients hospitalized with atrial fibrillation: results from the SAFETY trial

Background: The potential impact of disease management to optimize quality of care, health outcomes, and total healthcare costs across a range of cardiac disease states is unknown. Methods: A trial-based cost-utility analysis was conducted alongside a randomized controlled trial of 335 patients with chronic, non-valvular AF (without heart failure; the SAFETY Trial) discharged to home from three tertiary referral hospitals in Australia. A home-based disease management intervention (the SAFETY intervention) that involved community-based AF care including home visits was compared to routine primary healthcare and hospital outpatient follow-up (standard management). Bootstrapped incremental cost-utility ratios were computed based on quality-adjusted life-years (QALYs) and total healthcare costs. Cost-effectiveness acceptability curves were constructed to explore the probability of the SAFETY intervention being cost-effective. Sub-group analyses were performed based on age and sex to determine differential cost-effectiveness. Results: During median follow-up of 1.75 years, the SAFETY intervention was associated with a non-statistically significant increase in QALYs (0.02 per person) and lower total healthcare costs (–$4,375 per person). Although each of these findings were not statistically significant, the SAFETY intervention was found to be dominant (more effective and cost saving) in 58.8% of the bootstrapped iterations and cost-effective (more effective and gains in QALYs achieved at or below $50,000 per QALY gained) in 61.5% of the iterations. Males and those aged less than 78 years achieved greater gains in QALYs and savings in healthcare costs. The estimated value of perfect information in Australia (the monetized value of removing uncertainty in the cost-effectiveness results) was A$51 million, thus demonstrating the high potential gain from further research. Conclusions: Compared with standard management, the SAFETY intervention is potentially a dominant strategy for those with chronic, non-valvular AF. However, there would be substantial value in reducing the uncertainty in these estimates from further research.