Using Telemedicine to Detect Diabetic Retinopathy in American Indian/Alaska Natives and Other Ethnicities

Methods: We randomly selected diabetic patients to participate in a store-and-forward telemedicine program using non-mydriatic cameras (Camera group). We performed retinal photography at least once per year and used the International Diabetic Retinopathy scale to stage DR from Stage 1 (mild NPDR) to Stage 5 (proliferative DR). Diabetic educators and providers also encouraged the participants to have an annual eye exam with an eye care provider.

Results: We randomized 547 participants: 285 (52%) to the Camera group and 262 (48%) to the Control group. The Camera group had images evaluated in 224 (79%), with 14 (6.3%) having images that were unreviewable. Only 105 (40%) of the Provider group participants had an eye exam. The initial evaluation in the Camera group (more severe eye) showed 32 (14.2%) had mild NPDR; 7 (3.1%) were moderate; and 2 (0.9%) had very severe NPDR. Of those with images in the Camera group, DR worsened (> 1 stage) in 7 persons (3.0%); DR improved (< 1 stage) in 3 persons (1.3%); and stayed the same in 36 persons (16.1%).

Conclusion: Retinal imaging using a non-mydriatic camera increases the proportion of diabetics who obtain screening eye exams. Most diabetics in this sample did not have levels of diabetic retinopathy that required ophthalmic intervention.