Overview
The purpose of this study is to determine if use of a nonmydriatic fundus camera using autonomous artificial intelligence software at the point of care increases the proportion of underserved youth with diabetes screened for diabetic retinopathy, and to determine the diagnostic accuracy of the autonomous AI system in detecting diabetic retinopathy from retinal images of youth with diabetes.
Description
This study will recruit up to 500 individuals ages 8-21 with type 1 or type 2 diabetes. In this study, participants will undergo a point-of-care diabetic eye exam using autonomous AI software on a non-mydriatic fundus camera. Participants will receive the diabetic eye exam results immediately from the autonomous AI system, and if abnormal will be referred to an eye care provider for a dilated eye exam.
In the AI for ChildrenS Diabetic Eye ExamS Study (ACCESS2), 398 participants will be enrolled to determine if point of care autonomous AI increases the proportion of minority and underserved youth screened for diabetic retinopathy. The autonomous AI interpretation will also be compared to consensus grading of retinal specialists to determine if there is agreement and to determine the diagnostic accuracy of the system in youth.
A cohort of youth with known diabetic retinopathy (true positives) will also be enrolled as an enriched population to determine the diagnostic accuracy of autonomous AI compared to the prognostic standard interpretation of a central reading center.
Eligibility
Inclusion Criteria:
Meets American Diabetes Association (ADA) criteria for diabetic retinopathy screening:
- Diagnosis of Type 1 diabetes for ≥3 years, and age 11 or in puberty
- Diagnosis of Type 2 diabetes
Enriched cohort:
- Patients with Type 1 or Type 2 diabetes,
- 8-21 years of age with known diabetic retinopathy (true positives).
- No time limit on last diabetic eye exam.
Exclusion Criteria:
- Known diabetic eye exam in the last 12 months