Health economic simulation modeling of an AI-enabled clinical decision support system for coronary revascularization

While artificial intelligence (AI) models have been developed to support coronary revascularization decision-making, health economic evaluation of such models has been rare. We conducted a retrospective health economic simulation modeling study using real-world data from 25,942 adult patients with obstructive coronary artery disease in Alberta, Canada to evaluate the economic value of an AI-enabled coronary revascularization decision support system. Clinicians deciding among medical therapy only, percutaneous coronary intervention, and coronary artery bypass grafting were simulated to be provided with AI predictions of 3- and 5-year major adverse cardiovascular events and all-cause mortality. At a willingness-to-pay of $50,000 per quality adjusted life year (QALY), as many as 72.4% of all actual treatment decisions shifted to a different health economically optimized treatment, resulting in an average cost saving of $22,960 and a QALY gain equivalent to up to $22,439 per patient. Even in a conservative scenario where clinicians' AI adoption was assumed to be limited, 53.2% of the actual decisions shifted, resulting in an average QALY gain equivalent to up to $32,214 per patient. AI can potentially optimize the health system level economic value of treatment decisions in the form of reduced costs stemming from fewer future complications and improved patient outcomes.