Non-Ophthalmologist-Led Screening

Human grader-based screening at the health and wellness centre achieved high sensitivity and specificity for diabetic retinopathy detection. Human grader-based screening at the health and wellness centre had high sensitivity and specificity for referable diabetic retinopathy. The non-ophthalmologist-led health and wellness centre model performed better diagnostically and operationally than the home-based offline AI model. A trained non-ophthalmologist acquired fundus images after a structured 15-day training programme at a tertiary retina clinic. The health and wellness centre model supports task shifting in primary care but requires infrastructure, training and quality image acquisition.