Redesigning diabetic retinopathy (DR) screenings may provide mechanisms for a higher yield of disease by diagnosing more patients and a means of efficient referral to an ophthalmologist for treatment.1

With timely diagnosis and treatment, the dreaded complication of blindness associated with DR is preventable in >50% of cases. Access to screening or adherence to screening recommendations, however, is often problematic for patients. Many patients live in rural areas with poor access to eye care. Nonmydriatic fundus photography provides an effective mechanism for screening, particularly in these individuals.

The objective of our study was to demonstrate the feasibility and outcomes of a telemetric screening program using nonmydriatic fundus photography. Images were transmitted automatically from remote sources using the Internet to a reading center and then graded according to the degree of retinopathy found.

RECOMMENDATION FOR FOLLOW-UP
Depending on the readings, each patient received a recommendation regarding appropriate timing for follow-up with an ophthalmologist. Patients were also given a list of ophthalmologists in their area and who accepted their health plan.

We used the Topcon TRC-NW6S (Topcon Medical Systems, Inc., Paramus, NJ) nonmydriatic fundus camera that provides a viewing angle of 45º and produces nonstereoscopic fundus images (see sidebar). We recruited 589 adults with known diabetes from three sites in Pittsburgh; a general internal medicine clinic (site A, n=158), an endocrinology clinic (site B, n=110), and a mobile unit that traveled to screening events in communities (site C, n=321).

Each patient underwent nonmydriatic fundus photography of both eyes, and information on comorbidities, demographics, and medications were self-reported at the time of imaging. Images were transmitted telemetrically to a reading center and graded. Patients received a recommendation regarding follow-up care, and then investigators assessed patient compliance with the recommended follow-up plan by phone.

AGE AND COMORBIDITIES
The patients’ mean age was significantly higher at site C (mean, 61.4 years), as was the age at diagnosis of diabetes (mean 53.7 years). The comorbidities of hypertension (73.4%, P=.0001), neuropathy (37.9%, P=.0001), renal disease (12.0%, P=.0001), and depression (31.7%, P= 0.02) were significantly higher at site A. The majority of patients, 47.5%, 44.6%, and 52.8% had their last dilated eye exam >12 months prior for each site, respectively, A, B, and C.

Although it was recommended that the majority of patients see an eye doctor in 1 year (47.5%, 59.1%, and 67.3%, respectively), a small percentage of patients received a recommendation to see a doctor as soon as possible for immediate evaluation or treatment (1.3%, 1.8%, and 0.6%), respectively.

SUCCESSFUL IMPLEMENTATION IS NEEDED
We concluded that redesigning diabetic retinopathy screenings may provide mechanisms for a higher yield of disease and efficient referral to an ophthalmologist for treatment. Although numerous studies have been done to validate the use of single-field nonmydriatic fundus photography for DR screening, implementation of an effective large-scale screening program has not been successfully undertaken.

Successful implementation of a clinic- and community-based screening program could facilitate access for people with diabetes to screening eye exams and timely treatment of DR. The establishment of such a program will also enhance awareness of the importance of routine eye examinations among the diabetic population as well as the physicians who care for them.

Laura A. Bettencourt, MPH, is from the University of Pittsburgh Diabetes Institute. She may be reached at Bettencourtla3@upmc.edu; or phone: 412-692-4286.

1. Bettencourt LA, Zgibor J, Silowash R, et al. Outcomes from a diabetic retinopathy screening study implemented in clinic and community settings. Presented at the American Diabetes Association 67th Scientific Sessions. June 22-26, 2007. Chicago.