Study of Artificial Intelligence to Screen for Diabetic Retinopathy in Adults with Diabetes in Primary Care

What this Trial Is About

Researchers are evaluating the effects of universal screening for diabetic retinopathy DR and diabetic macular edema DME using artificial intelligence AI to interpret fundus photographs taken by trained nursing assistants with a portable fundus camera in a primary care setting. This study compares AI interpretation with images interpreted remotely by ophthalmologists to understand how AI might impact timely access to treatment for people with diabetes. Participants undergo mobile retinal photography performed by a trained nursing assistant under pupil dilation. Those in the AI group have their images analyzed by an AI program to determine if they have referable DR, while those in the ophthalmologist group have their images interpreted remotely by eye specialists. Based on these results, patients are classified for referral to a specialist for further care. During the study, researchers will track the number of appropriate referrals to ophthalmologists and monitor referrals for laser treatment or pharmacological intraocular treatments over about one year. They will also observe any referral failures for an average of 15 months after study completion. The study involves a randomized design with quadruple masking to ensure unbiased results.

Artificial Intelligence for Diagnosing Diabetic Retinopathy in Primary Care