Economic Evaluation of Artificial Intelligence Systems Versus Manual Screening for Diabetic Retinopathy in the United States

BACKGROUND: The objective of this economic model-ing study was to compare the cost effectiveness of fully automated retinal image screening (FARIS) to the cur-rent practice of universal ophthalmologist referral for diabetic retinopathy in the United States (US) health care system.METHODS: A Markov decision-analytic model was used to compare the automated versus manual screen-ing and management pathway for diabetic patients with unknown retinopathy status. Costs (in 2021 US dollars), quality-adjusted life year (QALY) gains, and incremental cost-effectiveness ratios were calculated. Sensitivity analysis was performed against a $50,000/ QALY willingness-to-pay threshold. RESULTS: FARIS was the dominant screening strategy, demonstrating cost savings of 18.8% at 5 years with equivalent net QALY gains to manual screening. Cost-effectiveness status was dependent on FARIS detec-tion specificity, with a threshold value of 54.8%.CONCLUSION: Artificial intelligence-based screening represents an economically advantageous screening modality for diabetic retinopathy in the US, offering equivalent long-term utility with significant poten-tial cost savings.