Diabetic Retinopathy Screening in Primary Care Real Practice: Study Procedures and Baseline Characteristics from the...

Area of Science:

• Ophthalmology
• Primary Care Medicine
• Artificial Intelligence

Background:

• Rising diabetes rates necessitate early detection of diabetic retinopathy (DR), a major cause of vision loss.
• Integrating DR screening into primary care settings shows promise for improved patient outcomes.
• Technological advancements, including AI, offer potential to enhance DR screening processes.

Purpose of the Study:

• To evaluate the agreement among primary care physicians (PCPs), ophthalmologists, and an AI system in DR screening.
• To assess the concordance of referral decisions for DR made by PCPs, ophthalmologists, and AI.
• To investigate the feasibility of implementing AI-assisted DR screening within a real-world primary care environment.

Main Methods:

• Retrospective collection of 1517 retinographies from 301 diabetic patients using a non-mydriatic retinal camera.
• Analysis of DR screening and referral decisions made by PCPs.
• Comparison of PCP decisions with ophthalmologist and AI-based assessments (planned for future analysis).

Main Results:

• Primary care physicians referred 34.5% of patients for ophthalmological evaluation.
• Referrals were primarily for suspected DR, retinal diseases, or media opacification.
• 13.62% of participants were suspected of having DR, with 9.63% receiving a confirmed diagnosis.

Conclusions:

• Initial findings highlight the role of PCPs in DR screening and referral.
• Further analysis will determine concordance between PCPs, AI, and ophthalmologists.
• Results are expected to support the integration of effective DR screening strategies into clinical practice.