Economic evaluation of a randomized controlled trial of a community-supported agriculture intervention for American Indians: the Go Healthy study

Abstract Background To conduct economic evaluations of a community-supported agriculture (CSA) dietary intervention tailored for American Indians (AIs). Methods A randomized controlled trial tested a 16-week multi-component CSA intervention in Oklahoma. We conducted cost-effectiveness, cost-utility, and cost-benefit analyses from a societal perspective: estimating the intervention’s marginal effect on HbA1c, converted HbA1c changes into quality-adjusted life years (QALYs), summing program, administrative, and participant costs, and calculating incremental cost-effectiveness and utility ratios (ICER, ICUR), plus non-health benefits, healthcare costs saved from averted diabetes, and net benefit (2024 USD). We addressed uncertainty with deterministic sensitivity analyses (alternative LY/QALY conversion methods; one-way variation in produce-box cost, wage rate, LY conversion factor, discount rate) and a probabilistic sensitivity analysis (Monte Carlo, 20,000 iterations) for the ICER and ICUR. The CBA used a $50,000–$150,000/QALY willingness-to-pay range and a conservative, age-specific excess-cost approach for healthcare savings from averted diabetes. Results Comparing the CSA treatment (n = 193) and control (n = 203) groups in balanced panel data, the 16-week intervention was cost-effective, with an ICER of $1,998.56 per 1% HbA1c reduction (one-way sensitivity range: $1,857.66–$2,139.04; probabilistic 95% simulation interval: $1,361–$3,587). Moreover, it yielded at least 8.88 QALYs (conservative base-case estimate; 15.33–16.02 QALYs under less conservative LY-conversion assumptions) from HbA1c reductions and an ICUR of $16,326.97 per QALY (sensitivity range: $9,050.16–$17,405.24; probabilistic 95% simulation interval: $6,139.28–$20,154.54), well below the willingness-to-pay (WTP) threshold of $50,000–$150,000 per QALY. Using this WTP range and a more conservative, age-specific excess-cost estimate for averted diabetes cases, the intervention would generate $444,000–$1,332,000 worth of QALYs, $79,183.13 in productivity gains, and $145,218.10 in healthcare costs saved from potentially averted diabetes cases, totaling $524,008.40–$1,412,008.40 in net benefit. Finally, assuming a 3% discount rate, net benefits of $13.10–$35.83 million and $16.95–$39.67 million would be gained over a decade, and $25.26–$69.06 million and $32.82–$76.62 million over a lifetime, with and without continuous intervention, respectively. Conclusion The indigenous CSA dietary intervention improved glycemic control and demonstrated short- and long-term cost-effectiveness relative to accepted willingness-to-pay thresholds, suggesting it may be a worthwhile investment for the AI population, subject to the limitations discussed herein and confirmation in larger and longer-term studies. Trial registration ClinicalTrials.gov ID NCT05773833. Registration date 2023-03-17.