Cost-effectiveness analysis of digital medicine support models for alcohol use disorder

Alcohol use disorder (AUD) is a major public health issue yet often under-addressed in primary care. Digital health interventions offer scalable support, but health systems require economic data to guide implementation decision-making. In a randomized clinical trial involving 558 participants registered at clinicaltrials.gov (NCT04011644; July 3, 2019), we compared three support models built on a shared digital medicine platform: self-monitored use (SM), peer-supported use (PS), and clinically integrated use (CI). We estimated delivery costs and effectiveness based on reductions in heavy drinking days. Annualized reductions in heavy drinking days were 42.4, 40.2, and 48.3 days per patient for SM, PS, and CI models, with per-patient costs of $162, $524, and $802, respectively. Average cost-effectiveness was $3.8 (SM), $13.0 (PS), and $16.6 (CI); PS was dominated, and the ICER for CI versus SM was $107.9. When downstream healthcare utilization was included, results suggested that more intensive support may be economically advantageous. Modeling implementation at scale showed that PS becomes cost-efficient for large patient populations. These findings highlight that no single model is universally optimal: health systems may benefit from offering multiple support levels to match patient preferences and organizational capacity.