Predicting Diabetic Retinopathy From Risk Factor Data and Digital Retinal Images

N/AActive Not RecruitingNCT03694145

Charles Drew University of Medicine and Science300 enrolled

Overview

The objective of this study is to compare the results of a deep learning approach to diabetic retinopathy assessment with results from (1) an in-person examination with an ophthalmologist, and (2) the assessments of optometrists involved in a teleretinal screening program.

This study represents the third aim of a grant with five aims. The study will compare and evaluate the predictive accuracy of: (a) machine learning models developed to grade diabetic retinopathy and assess the presence or absence of diabetic macular edema and (b) the assessments of optometrist readers, both from digital retinal images, against standard of care dilated retinal examinations by board-certified ophthalmologists and/or retinal-specialty fellows for 300 diabetic patients utilizing a Los Angeles County reading center. For the study, the investigators will recruit 300-500 eligible diabetic patients for in-person eye examinations performed by board certified ophthalmologists and/or retinal-specialty fellows at Los Angeles County reading centers. The study will take place over the course of two visits: a teleretinal screening and an in-person eye examination. The in-person dilated eye examinations that the study participants will participate in and be compensated for follow the usual standard of care that patients receive in a setting that does not utilize teleretinal screening. Yearly dilated eye examinations are standard of care for all persons with diabetes.

Study Type

OBSERVATIONAL

Enrollment

300

Conditions

Diabetic Retinopathy Diabetic Macular Edema

Interventions

In-Person Eye ExaminationOTHER

Dilated in-person eye examination by a board-certified ophthalmologist or retinal fellow.

Eligibility

Sex: ALLMin age: 18 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Patients diagnosed with Type I or Type II Diabetes * Patients who are over the age of 18 years * Patients who have not previously been seen in the current year by the LACDHS Teleretinal Diabetic Retinopathy Screening Program and Reading Center * Patients whose teleretinal screening exam images yield readable results Exclusion Criteria: * Patients under the age of 18 years * Patients with gestational diabetes * Patients who have previously been seen in the current year by LACDHS's Teleretinal Diabetic Retinopathy Screening Program and Reading Center * Patients whose teleretinal screening exam images do not yield readable results, as gradable images are needed for later comparison against ophthalmologist reads. * Previously eligible patients who do not return for an in-person eye exam within 3 months of receiving a teleretinal screening (In order for the results of the teleretinal screening and in-person eye examinations to yield similar information, patients who do not return for their in-person eye exam within 3 months of their teleretinal screening will not be able to remain in the study. This is because significant eye changes not documented by the teleretinal screening may occur after a 3-month period).

Locations (3)

Los Angeles Department of Public Health

Los Angeles, California, United States

University of California - Los Angeles

Los Angeles, California, United States

Charles R. Drew University of Medicine and Science

Los Angeles, California, United States

Outcomes

Primary Outcomes

Proportion of patients accurately diagnosed with retinopathy

Proportion of patients accurately diagnosed with retinopathy using machine learning versus proportion accurately diagnosed by teleretinal screening optometrists with in-person eye examinations by ophthalmologists used as a gold standard.

Time frame: 11/2022

Data from ClinicalTrials.gov

This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.