OBJECTIVES: Patients with atrial fibrillation (AF) who undergo percutaneous coronary intervention (PCI) are candidates for triple therapy with an anticoagulant and two antiplatelet agents. Triple therapy carries a significant risk of bleeding and several clinical trials have shown that dual therapy in which one antiplatelet agent is omitted carries significantly lower bleeding risk. In this study, we used a Markov model to assess the cost-effectiveness of dual versus triple therapy using data from the recent AUGUSTUS trial.

METHODS: We created a Markov model with two treatment arms: dual therapy (anticoagulant + P2Y12 inhibitor) and triple therapy (anticoagulant + P2Y12 inhibitor + aspirin). We used event rates from the AUGUSTUS trial and cost and utility estimates from the medical literature. The following health states were modeled: well with AF (baseline), myocardial infarction, minor ischemic stroke, major ischemic stroke, intracranial hemorrhage, minor bleeding, major bleeding, and death. Patients were 70-years-old with AF and recent PCI and were followed for a lifetime horizon from the US healthcare system perspective. Outcomes included quality-adjusted life-years (QALYs), costs (2019 USD), and incremental cost-effectiveness ratios (ICERs).

RESULTS: Triple therapy resulted in total costs of $305,304 and 14.64 QALYs and dual therapy resulted in total costs of $347,208 and 14.98 QALYs. The ICER for dual therapy was $122,143/QALY. In one-way sensitivity analyses, the model was most sensitive to intracranial hemorrhage and ischemic stroke event rates. Probabilistic sensitivity analyses with Monte Carlo simulations demonstrated that dual therapy was cost-effective 54% of the time at a willingness-to-pay (WTP) threshold of $100,000/QALY and 61% of the time at a WTP of $150,000/QALY. Triple therapy was dominant in 26% of simulations and dual therapy was dominant in 12% of simulations, indicating uncertainty within the model.

CONCLUSIONS: Dual therapy is likely cost-effective compared to triple therapy albeit with a degree of uncertainty.