<![CDATA[Introduction: Diabetic retinopathy (DR) is a leading cause of preventable blindness, yet screening rates remain critically low in rural and underserved settings due to limited access to ophthalmologists and logistical barriers. Tele-ophthalmology and artificial intelligence (AI)-based retinal imaging have emerged as potential solutions, but their effectiveness in rural contexts requires systematic evaluation. Methods: This systematic review synthesized evidence from 38 studies, including randomized controlled trials, etc. Studies were included if they evaluated tele-ophthalmology or AI-based DR screening in rural, remote, or underserved populations and reported outcomes related to diagnostic accuracy, screening completion, referral adherence, cost-effectiveness, or implementation factors. Results: Tele-ophthalmology significantly improved screening completion rates, with RCT demonstrating 94% completion in the telemedicine arm versus 56% in traditional surveillance (p<0.001). Referral adherence varied substantially from 13-52% depending on financial and logistical support. Pooled diagnostic sensitivity for referable DR ranged from 0.90-0.95 across tele-ophthalmology and AI modalities, with specificity of 0.81-0.91. Cost-effectiveness analyses consistently favored tele-ophthalmology, with ICERs ranging from cost-saving to $3,328 per QALY in rural settings. Discussion: Tele-ophthalmology effectively addresses the access gap in rural DR screening. AI-based systems offer comparable diagnostic accuracy to human graders with added advantages in speed and immediate feedback, which can improve referral adherence when combined with patient support. However, long-term screening participation tends to decline without sustained engagement strategies. Implementation success depends on reliable infrastructure, task-shifting to non-ophthalmologist staff, and integration with functional referral pathways. Conclusion: Tele-ophthalmology and AI-based screening are effective, accurate, and cost-effective for DR screening in rural settings. Successful programs require concurrent investment in referral infrastructure, patient navigation support, and quality assurance mechanisms. Future implementations should prioritize context-specific validation and sustainable engagement strategies.]]>
Copyrights © 2026
