Clinical Decision Support Algorithm to Predict Diabetic Retinopathy

Validating a Clinical Decision Support Algorithm Developed With Demographic, Co-morbidity, and Lab Data to Diagnose, Stage, Prevent, and Monitor a Patient's Diabetic Retinopathy

Diabetic retinopathy (DR), a complication of diabetes, is a leading cause of blindness among working-aged adults globally. In its early stages, DR is symptomless, and can only be detected by an annual eye exam. Once the disease has progressed to the point where vision loss has occurred, the damage is irreversible. Consequently, early detection is quintessential in treating DR. Two barriers to early detection are poor patient compliance with the annual exam and lack of access to specialists in rural areas. This research is focused on developing and validating new, cost-effective predictive technologies that can improve early screening of DR. Our overall objective is to develop and implement an entire suite of tools to detect diabetes complications in order to augment care for underserved rural populations in the US and internationally.

Inclusion Criteria:

For diabetic patients with DR:

• With 250.xx diabetes and 362.0x DR ICD-9 codes
• All variables are complete within the observation window

For diabetic patients without DR:

• With 250.xx diabetes ICD-9 codes
• Without 362.0x DR ICD-9 codes
• All variables are complete within the observation window