AACE 2026: AI-assisted retinal screening expands diabetes eye care
Artificial intelligence-assisted retinal imaging embedded within primary care and endocrinology workflows more than doubled diabetic retinopathy screening rates among patients with diabetes during a 12-month quality improvement initiative, according to research presented at the 2026 AACE Annual Meeting in Las Vegas.
Screening rates increased from 8% before the intervention to 18% afterward, representing an 118% relative increase.
Artificial intelligence (AI) was integrated into a broader screening initiative called Saving Sight, which also combined continuing medical education, electronic health record integration, patient outreach, and coordinated referral management.
For the AI component, retinal images captured through digital retinal cameras underwent AI-supported image review using deep learning systems. Patients with positive screenings were then referred for ophthalmologic follow-up. The approach allowed retinal screening to occur directly within primary care and endocrinology clinics rather than relying exclusively on ophthalmology referrals.
The AI-assisted screening program identified 1,023 patients at risk for diabetic macular edema among 5,396 patients examined during the 12-month initiative.
Additional findings included the following:
Among 18,729 eligible patients with diabetes before the 12-month intervention and 18,272 afterward, the total number screened increased from 2,600 to 4,398 patients.
Referrals for screening increased from 2,313 before the intervention to 3,092 afterward.
Electronic health record and claims analyses demonstrated a 27% increase in screening rates across participating groups and a 23% increase in retinal referrals.
In addition to screening outcomes, the program reported improvements in clinician education and practice behaviors. The initiative was associated with a 77% improvement in explaining diabetic retinopathy risk factors and screening importance, a 67% increase in describing treatment options, and a 67% improvement in recommending anti-vascular endothelial growth factor therapy for patients with diabetes.
The study was conducted by investigators affiliated with DKBmed, New York. Funding sources, disclosures, or conflicts of interest were not reported.
AACE Endocrine AI is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE®). The ideas and opinions expressed in AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies.