Abstract
Purpose: :
To evaluate the efficacy of diabetic retinopathy screenings at Alameda County Medical Centers.
Methods: :
A digital fundus camera (Canon CR6-45NM) was used to screen 102 patients (48 females/54 males) who were at high-risk for sight-threatening diabetic retinopathy. Patients had an average age of 54.7 ± 9.5 years and their mean duration of diabetes was 10.1 ± 6.4 years. On average, patients reported having had an eye exam within the last 2.0 ± 1.7 years. Patients were comprised of four main ethnicities: Hispanic (47.1%), African American (22.5%), Asian (18.6%), and Caucasian (5.9%).
Results: :
The diabetic retinopathy screening referred 45 patients (22 had severe non-proliferative diabetic retinopathy, 29 had clinically significant macular edema, and 11 had proliferative diabetic retinopathy). A dilated fundus examination performed after the screening (6.6 ± 7.3 weeks later) determined 46 patients had referable retinopathy (21 had severe non-proliferative diabetic retinopathy, 29 had clinically significant macular edema, and 14 had proliferative diabetic retinopathy). The sensitivity and specificity of the screening, compared to the dilated fundus examination was 81.5% and 87.8% for clinically significant macular edema and 71.4% and 98.9% for proliferative diabetic retinopathy. Multivariate analysis revealed referrals from the dilated fundus examination were associated with older age (P=0.003), lower visual acuity (P=0.02), and longer duration of diabetes (P=0.04).Of the patients who were referred to ophthalmology from the dilated fundus examination, 27 (58.7%) were subsequently examined by that department. The average time period between the dilated fundus examination and an ophthalmology appointment was 139.0 ± 82.7 days (more than 4 months). Multiple regression analysis did not find any variable (including an ophthalmology appointment within 90 days) that was significantly associated with a patient missing or attending an ophthalmology appointment.
Conclusions: :
Diabetic retinopathy screenings at Alameda County Medical Centers detected sight-threatening retinopathy with good sensitivity and specificity. To be more effective, they should consider targeting older patients with longer duration of diabetes. The implementation and operation of diabetic retinopathy screenings, in general, require constant evaluation and refinement to be useful. Additional studies that follow patients referred from diabetic retinopathy screenings are required.
Keywords: diabetic retinopathy • detection • diabetes