Related Experiment Videos
Methods to quantify the relation between disease progression in paired eyes.
1Brigham and Women's Hospital and Harvard Medical School, and Department of Biostatistics, Harvard School of Public Health, Boston, MA, USA. rglynn@rics.bwh.harverd.edu
American Journal of Epidemiology
|June 15, 2000
Summary
Diabetic retinopathy progression in one eye significantly increases the risk of progression in the other eye. Joint survival models provide more stable estimates of this inter-eye risk for diabetic retinopathy.
Area of Science:
- Ophthalmology
- Biostatistics
- Endocrinology
Background:
- Diabetic retinopathy (DR) is a leading cause of vision loss in diabetic patients.
- Understanding the inter-eye relationship in DR progression is crucial for risk assessment and management.
Purpose of the Study:
- To compare methods for quantifying the increased risk of DR progression in one eye following progression in the fellow eye.
- To evaluate the utility of joint survival models for analyzing correlated eye progression data.
Main Methods:
- Analysis of data from 478 patients with insulin-dependent diabetes mellitus from the Sorbinil Retinopathy Trial (1983-1988).
- Utilized eye-specific proportional hazards models and two alternative joint proportional hazards models accounting for inter-eye correlation.
- Follow-up duration was a median of 41 months.
Main Results:
- Crude incidence rates showed significantly higher DR progression in one eye after the fellow eye progressed (7.7x for right eyes, 4.4x for left eyes).
- Eye-specific models yielded relative risks of 2.6 for right eyes and 1.4 for left eyes.
- Joint survival models estimated relative risks of 1.9 and 2.7, indicating a significant inter-eye risk.
Conclusions:
- Progression of diabetic retinopathy in one eye is a significant risk factor for progression in the other eye.
- Joint survival models integrating data from both eyes offer more robust and stable estimates of this inter-eye risk compared to eye-specific analyses.