Rapid Results in DR Screenings Improve Patient Follow-up Adherence

Adding artificial intelligence tools to retinal cameras could reduce inequality in access to DR screening, but its cost-effectiveness may depend on individual countries’ wealth and AI infrastructure. Photo:  Thomas A. Wong, OD. [Epub ahead of print]. Click image to enlarge. Catching diabetic retinopathy (DR) as early as possible is key to preserving patients’ vision. But despite the frequent implementation of large-scale screening efforts, adherence to annual eye exams still lag, with rates presently at 67% in high-income countries and 39% in low- and middle-income countries, according to research. Those patients with referable DR have also been slow to attend referral appointments. To probe for ways to improve adherence, researchers recently compared human graders and computer-based AI assessment of images in DR screening programs. Their review study published recently in Eye showed that AI algorithms are associated with an increased uptake in follow-up exams because they give patients actionable information right away.The researchers screened articles in several major health sciences databases. Their review yielded data from 20,108 patients with diabetes across six studies, 6,476 of which were graded using AI and 13,632 who were graded by humans. The random effects model showed that initial AI assessment for diabetic retinopathy significantly increased follow-up appointments (OR 1.89). Patients under 21 in particular demonstrated higher uptake (OR 11.06) compared with adults (OR 2.75).“AI-assisted DR screening allows real-time classification of severity of DR at the point-of-care, in any clinical setting, irrespective of availability of skilled ophthalmic personnel; this in turn allows patients to be informed of any need for onward referral sooner than can be achieved with human graders,” the researchers explained in their Eye paper. “AI based DR screening systems accelerate the analysis of fundus images and provide DR grading results faster than human grader-based systems.”The researchers concluded that this increased uptake in follow-up exams is “likely due to instant results being made available with AI-based algorithms when compared to a delay in the communication assessment outcomes achieved with human graders.”Click here for the journal source. Rahmati M, Smith L, Prabhath M, et al. Artificial intelligence improves follow-up appointment uptake for diabetic retinal assessment: a systematic review and meta-analysis. Eye 2025. [Epub May 30, 2025].