Diabetes is inherently challenging to treat as it is so complex and data-intensive. As this patient population grows and our offices get busier, the challenge will only become more acute. The field of diabetes has been revolutionized in not only how we treat our patients, but also in the way that we are able to diagnose and monitor them. As an endocrinologist who has embraced technology and its advances over the years, I am impressed by how underutilized that the available tools are. This article will review the evolving role of technology in caring for our patients, specifically in regard to diabetic eye care.

SCREENING FOR DIABETIC RETINOPATHY

Screening for diabetic retinopathy is an essential task for those who manage patients with diabetes. Diabetic retinopathy has been shown to be a prevalent comorbidity and to be the complication that has the greatest impact on quality of life.1-3 If patients are properly screened for retinopathy they can be referred to the ophthalmologist earlier, potentially preventing vision loss.

Who should be screened? According to the American Diabetes Association, patients with Type 1 diabetes should be screened for diabetic retinopathy after 10 years with the disease within 5 years of diagnosis.4 Patients with Type 2 diabetes, however, should be screened upon diagnosis and yearly thereafter. It is reasonable to consider screening at 2-3 year intervals for patients who are considered normal after initial or subsequent screening. For our patients who are pregnant, we should be screening them for diabetic retinopathy both before and during pregnancy, as retinopathy can exacerbate during this time.

The gold standard of screening for diabetic retinopathy are the guidelines set forth by the Early Treatment Diabetic Retinopathy Study (ETDRS), which can be seen in the sidebar “American Diabetes Association Guidelines for Screening for Diabetic Retinopathy.”

For various reasons, it may be difficult to have all of our patients screened by an ophthalmologist with a full retina exam. Figure 1 shows Medi-Cal data for 2010 on how screening for residents of California compare to the national average, indicating missed opportunities for screening; similar data are available for other states.

The barriers to universal screening for diabetic retinopathy include geography, economic status, social limitations, and lack of education regarding the importance of screening. Another barrier is the magnitude of the task of screening. There are approximately 160 000 ophthalmologists worldwide as of 2010 and an estimated 285 million patients with diabetes. This translates into 1781 patients per ophthalmologist. These numbers are projected to become less favorable in the future, with the number of patients with diabetes expected to balloon to 438 million by 2030 and the number of ophthalmologists to increase only to 224 000—1955 patients per every ophthalmologist.

Even if there were enough ophthalmologists to see every patient with diabetes, it is a challenge to persuade patients to see an eye doctor. It may be the required copay, or the distance, or simply the inconvenience of having to see another doctor. The first step that we can take as diabetologists is to learn how to the technology that will enable us to screen for this disease more effectively.

TECHNOLOGY FOR SCREENING

When deciding what kind of technology to include in your practice for screening patients, it is important to consider staffing and training to properly use the equipment. The gold standard to evaluate diabetic retinopathy is the ETDRS 30º 7-field color reference (Figure 2 with color fundus photography. A skilled photographer is necessary for this technology, and in my experience, such an individual can be harder to come by than an ophthalmologist. The most commonly employed technique to evaluate diabetic retinopathy is direct ophthalmoscopy. However, the literature shows that this to be an insensitive and inaccurate methodology, particularly to detect macular edema.5

An important advance to image the retina and screen for diabetic retinopathy is the nonmydriatic retina camera. This device does not require the patient to be dilated, making it more convenient for the patient and faster to use for the technician. It is also much lower in cost and is appropriate for a diabetes clinic. A new technology, 200º laser scanning ophthalmoscopy, may prove useful, but in my opinion, it must prove its worth before finding a “prime time” spot in the diabetologist's office, as this is a costly and more complicated device.

TELEMEDICINE

Image acquisition is only the first step in the process of retinal screening. The next step is the processing and interpretation of the images, which is where telemedicine potentially comes in. Because it is crucial to involve an ophthalmologist in the reading of the images to accurately screen for diabetic retinopathy, teleophthalmology may make this more cost-effective. A typical scenario is where the images are captured and then sent to an ophthalmology reading center to determine if the patient is a suspect for diabetic retinopathy.

The American Telemedicine Association has established categories of clinical validation for diabetic retinal performance in telemedicine6

Category 1. Category 1 validation indicates a system can separate patients into 2 categories: (1) those who have no or very mild nonproliferative diabetic retinopathy (ETDRS level 20 or below), and (2) those with levels of diabetic retinopathy more severe than ETDRS level 20. Functionally, Category 1 validation allows identification of patients who have no or minimal diabetic retinopathy and those who have more than minimal diabetic retinopathy.

Category 2. Category 2 validation indicates a system can accurately determine if sight-threatening diabetic retinopathy is present or not present as evidenced by any level of DME, severe or worse levels of nonproliferative diabetic retinopathy (ETDRS level 53 or worse), or proliferative diabetic retinopathy (ETDRS level 61 or worse). Category 2 validation allows identification of patients who do not have sight-threatening diabetic retinopathy and those who have potentially sight-threatening diabetic retinopathy. Patients with sight-threatening diabetic retinopathy generally require prompt referral for management.

Category 3. Category 3 validation indicates a system can identify ETDRS defined levels of nonproliferative diabetic retinopathy (mild, moderate, or severe), proliferative diabetic retinopathy (early, high-risk), and DME with accuracy sufficient to determine appropriate follow-up and treatment strategies. Category 3 validation allows patient management to match clinical recommendations based on clinical retinal examination through dilated pupils.

Category 4. Category 4 validation indicates a system matches or exceeds the ability of ETDRS photos to identify lesions of diabetic retinopathy to determine levels of diabetic retinopathy and DME. Functionally, Category 4 validation indicates a program can replace ETDRS photos in any clinical or research program.

The Joslin Vision Network is a Category 3 system, which is human-enabled. These types of systems are the most common in the United States and work well with defined populations such as the Veterans Health Administration and the Indian Health Service. Category 3 systems, however, are typically not utilized outside of these large systems or in private practice, due to the enormous startup and maintenance costs.

Figure 3 shows the workflow that has been used by a telemedicine company (Inoveon Corporation, Oklahoma City, OK) that is representative of what might be used by private practices and academic centers. It is a comprehensive system of getting the patients in, acquiring good images, and then moving the images to a reading center.

In my opinion, the future of screening technology will be in having a computerized method with which to identify potential abnormalities and flag them for a screener, making the process faster and less expensive. DrishtiCare, which is a telescreening platform that is in development in India, uses a server-based prescreening system to evaluate fundus images and then refer suspect images to human examiners.7 The workflow and analysis diagram for this program is seen in Figure 4. Critical to the success of all of these approaches is appropriate reimbursement for the service.

SUMMARY

The technology for point-of-care retinal examination is approaching the point where we will be able efficiently and affordably screen our patients with diabetes for retinopathy. This will allow for interventions that may drastically reduce the numbers of patients who suffer severe vision loss. With a point-of-care system, we would be able to reduce the amount of travel time and lost work hours for patients, ensuring that they are not lost to follow-up because of inconvenience. We could offer expert care in any geographic location using technology, such as nonmydriatic cameras, that are easier for both the technician and the patient. Additionally, we will be able to utilize automated systems that increase efficiency and reduce staffing costs. The drawbacks to telemedicine include a high initial capital cost, reimbursement issues, and the limitations of diagnosis for non-diabetic retinopathy pathologies. Despite the drawbacks and potential challenges, creating a system that will overcome the barriers to screening is a worthwhile task. This has the potential to increase our chances of getting patients screened with appropriate imaging and referred for treatment that has been demonstrated to reduce the overall incidence of vision loss and blindness from diabetes.

Timothy S. Bailey, MD, FACP, FACE, is an Endocrinologist and a Clinical Associate Professor at the University of California, San Diego School of Medicine. He is a Staff Physician at the Palomar-Pomerado Health System and director of AMCR Institute in Escondido California.

  1. Brown MM, Brown GC, Sharma S, Busbee B. Quality of life associated with visual loss: a time tradeoff utility analysis comparison with medical health states. Ophthalmology. 2003;110(6):1076-1081.
  2. Brown MM, Brown GC, Sharma S, Shah G. Utility values and diabetic retinopathy. Am J Ophthalmol. 1999;128(3):324-330.
  3. Brown MM, Brown GC, Sharma S, Kistler J, Brown H. Utility values associated with blindness in an adult population. Br J Ophthalmol. 2001;85(3):327-331.
  4. American Diabetes Association. Standards of medical care in diabetes: 2011. Diabetes Care. 2011;34(Suppl 1):S11-61.
  5. Hutchinson A, McIntosh A, Peters J, et al. Effectiveness of screening and monitoring tests for diabetic retinopathy: a systematic review. Diabet Med. 2000;17:495-506.
  6. American Telemedicine Association. Telehealth practice recommendations for diabetic retinopathy. February 2011. Available at: http://www.americantelemed.org/files/public/standards/ DiabeticRetinopathy_withCOVER.pdf.
  7. Joshi GD, Sivaswamy J. DrishtiCare: a telescreening platform for diabetic retinopathy powered with fundus image analysis. J Diabetes Sci Technol. 2011;5(1):23-31.