Context: Annual eye screening for patients with diabetes mellitus is frequently proposed as a measure of quality of care. However, the benefit of annual vs less frequent screening intervals has not been well evaluated, especially for low-risk patients.
Objective: To examine the marginal cost-effectiveness of various screening intervals for eye disease in patients with type 2 diabetes, stratified by age and level of glycemic control.
Design: Markov cost-effectiveness model.
Setting and participants: Hypothetical patients based on the US population of diabetic patients older than 40 years from the Third National Health and Nutrition Examination Survey.
Main outcome measures: Patient time spent blind, quality-adjusted life-years (QALYs), and costs of annual vs less frequent screening compared by age and level of hemoglobin A1c.
Results: Retinal screening in patients with type 2 diabetes is an effective intervention; however, the risk reduction varies dramatically by age and level of glycemic control. On average, a high-risk patient who is aged 45 years and has a hemoglobin A1c level of 11% gains 21 days of sight when screened annually as opposed to every third year, while a low-risk patient who is aged 65 years and has a hemoglobin A1c level of 7% gains an average of 3 days of sight. The marginal cost-effectiveness of screening annually vs every other year also varies; patients in the high-risk group cost an additional $40530 per QALY gained, while those in the low-risk group cost an additional $211570 per QALY gained. In the US population, retinal screening annually vs every other year for patients with type 2 diabetes costs $107510 per QALY gained, while screening every other year vs every third year costs $49760 per QALY gained.
Conclusions: Annual retinal screening for all patients with type 2 diabetes without previously detected retinopathy may not be warranted on the basis of cost-effectiveness, and tailoring recommendations to individual circumstances may be preferable. Organizations evaluating quality of care should consider costs and benefits carefully before setting universal standards.