Document resource
This scoping review examines the existing literature on the application of artificial intelligence (AI) in screening for eye diseases, with a focus on evaluating whether AI-assisted diagnostic technologies enhance the availability, accessibility, acceptability and quality of screening services. 42 original studies were selected for in-depth analysis, including those employing health economic evaluations. Methodological quality was assessed using the Mixed Methods Appraisal Tool, with the majority of studies demonstrating high quality—24 scored 5/5, 15 scored 4/5 and the rest scored 3/5. Among the included studies, 34 compared manual screening with either AI-assisted or fully AI-driven approaches. Availability was the most frequently studied aspect (28 studies), followed by acceptability (12 studies), whereas accessibility and service quality were less commonly addressed. Overall, AI shows significant potential to improve the cost-effectiveness of eye care services and enhance patient access, particularly in remote or underserved regions. It was also well-accepted by patients, with high satisfaction and improved referral compliance. The findings suggest that AI holds promise for advancing eye disease screening, although large-scale, long-term trials are needed to effectively integrate AI into the reconstruction of screening processes and the reshaping of eye health service systems.