[Skip to Content]
[Skip to Content Landing]
July 26, 2000

How Often Should Patients With Diabetes Be Screened for Retinopathy?

Author Affiliations

Stephen J.LurieMD, PhD, Contributing EditorIndividualAuthor

JAMA. 2000;284(4):437-439. doi:10.1001/jama.284.4.436

To the Editor: I would like to raise the following concerns about the study by Dr Vijan and colleagues.1 (1) The conclusions are based on modeling, not on a clinical study. (2) A lower utility value for blindness results in greater cost-effectiveness of annual screening. (3) Visual impairment other than blindness can cause disability, inability to drive, loss of work days, and decreased quality of life. Blurred vision has a significant impact on functional status and well-being.2 (4) Compliance with annual examinations and follow-up to monitor referrals is already low. If examinations are advised every 2 to 3 years, and patients miss those, it could be 4 to 6 years before some patients are seen. Diabetes remains the leading cause of blindness among working-age Americans. (5) Factors other than glycohemoglobin, such as duration of diabetes, use of insulin, sudden tight diabetic control,3 hypertension, lipid levels,4 and renal disease, are associated with risk of progression to blindness. (6) Nearly everyone with diabetes develops retinopathy. For patients with type 2 diabetes taking insulin and not taking insulin, retinopathy is found in 84% and 53% of patients, respectively, with up to 19 years of follow-up. Klein et al5 state that "Our data suggest that even persons with no retinopathy at baseline are in need of ophthalmologic observation because of the significant number at risk of progressing proliferative retinopathy during the 14 years of follow-up. In addition, our data confirm the current guidelines for ophthalmologic care for detecting proliferative retinopathy and clinically significant macular edema over the long-term course of diabetes." (7) The UK Prospective Diabetes Study (UKPDS) may not be representative of the United States. Under excellent care, patients had good glucose and blood pressure control, resulting in lower rates of progression. Potential differing risks of retinopathy in African Americans, Hispanic Americans, Native Americans, and Asian Americans were not considered. (8) Other reasons for examination include reinforcing the need for glucose and blood pressure control and detecting complications such as kidney and cardiovascular disease and eye conditions associated with diabetes. (9) Outside the model, all patients are not seen by an ophthalmologist. The real-world accuracy of detection is lower than modeled, thus reducing the effectiveness of screening.