feature
Improving Diabetic Eye Care
Telediagnosis studies show promise for this large patient population.
BY MICHAEL D. ABRÁMOFF, M.D., PH.D., AND STEPHEN R. RUSSELL, M.D.

Because 10% or less of patients may have diabetic retinopathy, retinas may be normal 90% of the time.

With more than 18 million people with diabetes in the United States as of 2005, recommended regular retinal examination of these patients may make up to 30% or even 40% of all visits in your practice. Because 10% or less of these patients may have diabetic retinopathy, you may be examining normal retinas 90% of the time.

Well-established guidelines for regular eye examinations have been recommended by the American Diabetes Association and the American Academy of Ophthalmology. Despite this, about 50% of patients with diabetes do not comply with these guidelines. Providing timely and effective treatment, enhanced by improved detection and early intervention, has been shown to reduce the incidence of diabetic visual loss, which is the most common cause of visual loss in patients of working age in the United States._¹ Efforts are underway on the local, state and national levels to increase compliance with these guidelines.

Ample evidence exists that regular dilated retinal exams, combined with appropriate intervention, reduce visual loss and blindness in patients with diabetes. But clearly, if you, as an ophthalmologist, are seeing only 50% of the people with diabetes, whatever treatment may be available is never going to reach the entire at-risk population._² Part of the explanation for this lack of compliance is that patients, family physicians and internal medicine specialists need to be reminded of the importance of regular eye exams. But a major factor is simply lack of accessibility of the retinal exam because of the inconvenience of traveling long distances to see an ophthalmologist, as well as the cost of the visit. Additionally, these patients often have no visual complaints.

Over the past 5 years, we at the University of Iowa have been working to make regular retinal exams for patients with diabetes more available. We maintain the quality of retinopathy detection through the Rural Iowa Telediagnosis project. Here, we will focus on how this project works and how projects such as this may affect the management of patients with diabetes in the future.

Telediagnosis and the Rural Iowa Telediagnosis Project

By leveraging the Internet, we examine the retinal photographs of patients at a distance. Family physicians who join the project have a retinal fundus camera that produces images of sufficient quality (more on that to follow). The fundus camera is located in a darkened room. At the start of each day, a nurse, tech or physician assistant logs onto the secure Web site to look at the diagnoses of the photographs taken the previous days. Then, for each patient with diabetes who visits the clinic for their bi-annual diabetes follow-up, the staff member enters relevant data including name, date of birth and serum HbA1C%. If the clinic has EMR, the data are pulled from the patient's record. Photographs are taken, and the clinical data and photographs are submitted through the secure Web site. We currently have cameras in 11 clinics across Iowa, and anticipate to be installing at least seven more this year in Iowa, Illinois and Nebraska, predominantly in rural or underserved areas.

At the University of Iowa, the retinal specialists log on to the same secure Web site every morning before clinic, and go through the patients' images one by one. The easy-to-use Web interface allows the examiner to then make a decision as to whether the image demonstrates abnormalities of diabetic retinopathy or any retinal disease by clicking on diagnostic choices.

Shortly thereafter, when a nurse or technician has logged on at the family physician's office (often more than 300 miles away), the diagnosis is printed in the form of a letter, or, if the physician has EMR, entered into the notes automatically. The photographs are accessible to patients, or their ophthalmologist, through an encoded Web site. If abnormalities are found, the family physician is advised to refer the patient to their ophthalmologist. Issues regarding referral are best handled among the providers affected. In addition, a number of hurdles exist, including concerns about quality, changing technology, legal ramifications and financial reimbursement.

Quality

Cameras ready to be shipped to remote locations for the Rural Iowa Telediagnosis project. From left to right: Dr. Abràmoff, Terry J. Protextor, executive director of University of Iowa Community Medical Services, Inc. and Keith Carter, M.D., head of the Department of Ophthalmology and Visual Sciences at the University of Iowa.

Can a retinal camera and the Internet replace a full retinal exam by an ophthalmologist? No, they cannot. Some cases of diabetic retinopathy may be missed using this approach. It is important to understand that the nature of the retinal exam, whether by indirect ophthalmoscopy or by retinal camera, makes it possible for diagnoses to be missed. It is probable
that more treatable lesions will be missed with a retinal
camera and the Internet, than by a retinal exam in the office.

Because the target population of this project are patients who would otherwise not be examined at all, a balance had to be found between quality and effort on the one, and cost on the other hand.

In addition to Rural Iowa Telediagnosis, other non-profit and for-profit diabetic retinopathy diagnostic projects exist, such as Inoveon (www.inoveon.com), EyeTel Imaging (www.eyetel-imaging.com), Joslin Vision Network (www.joslin.org/joslin_vision_network.asp) and Vanderbilt University's Ophthalmic Imaging Center (www.retinopathyscreening.org). Each has struck a different balance between quality and cost. The American Telemedicine Association (ATA) has set standards for telemedicine applications for diagnosis and classification of diabetic retinopathy.3 Because our ophthalmologist readers determine whether the images demonstrate diabetic retinopathy, and do not classify the severity or type of retinopathy, Rural Iowa Telediagnosis falls under the category II ATA. This approach has been shown to be cost effective.4

Technology

Only with the advent of broadband Internet and high-speed computers can the images of a patient be transmitted securely, allaying privacy concerns. The Web site interface must be easy to learn and use in order to diminish the number of errors, and to allow ophthalmologist readers to use ancillary diagnostic algorithms as they become available.

The Rural Iowa Telediagnosis technology was originally developed in 2000-2001 by Michael Abràmoff, M.D., Ph.D., when he was on the faculty at the Vrije Universiteit Amsterdam, the Netherlands, and has been continuously refined over the years to interface with generally available EMR, primarily DICOM compatible systems. An offshoot of this project, the EyeCheck project in the Netherlands, currently services more than 15,000 patients per year.5 Both the Rural Iowa Telediagnosis and EyeCheck projects require the use of high quality non-mydriatic cameras, such as the Topcon TRC-NW200 (Topcon America Corporation, Paramus, N.J.), the Kowa Non-Myd a-D Fundus (Kowa Optimed, Torrance, Calif.), the Canon CR-DGi (Canon U.S.A., Lake Success, N.Y.) and the VISUCAM Pro NM (Carl Zeiss Meditec, Dublin, Calif.). It is important to note that these cameras are used remotely, so quality paired with ease of use is essential. The price of such a camera may be prohibitive to many family physicians who may use it on only 200 or 300 patients per year. Thus far, we have managed cost issues either through grants from foundations or through large health insurance carriers such as Blue Cross/Blue Shield. A few patients do need to be dilated pharmacologically, but studies have shown that this is a small fraction of the total number.5

Liability/Reimbursement

Liability and reimbursement are also challenges with telediagnosis. The rules on Internet transmission are not explicitly outlined in the HIPAA guidelines, so we have chosen the highest security standard that is available over the Internet, namely SSL (Secure Socket Layer) which is also used for online banking. HIPAA probably does not allow any identifiable patient data to be transmitted over e-mail or non-secure Internet, as that would amount to copying the patient's data and giving it to anyone who asks.

Finally, and possibly the most important issue that can make or break projects such as this, is reimbursement. Every time we present aspects of the Rural Iowa Telediagnosis Project, the first question is about reimbursement. There currently is no CPT code for telediagnosis. There is a code for fundus photography, and so far we have been able to work out mutually satisfactory solutions with insurers such as Blue Cross/Blue Shield and CMS.

A technician adjusts the camera while a patient sits for a fundus photo

The Future

The Rural Iowa Telediagnosis project began 18 months ago and its growth is accelerating. One focus of our research is to provide computer-assisted diagnosis. In the near future we expect to start studying larger populations utilizing the retinal images read by a computer algorithm compared to a retinal specialist. Telediagnosis may then become cost effective for the early diagnosis of other eye diseases such as glaucoma.

Anecdotal evidence suggests that ophthalmologists who have their offices in an area where telediagnosis has been implemented often see their numbers of referrals increase, because a larger proportion of diabetes patients are diagnosed and referred in a timely way. We have recently received a small grant to study this question.

In summary, the future looks bright for telediagnosis in ophthalmology, and we hope this project can contribute to our nation's ophthalmologists striving to improve the sight of all Americans with diabetes.

Michael D. Abràmoff, M.D., Ph.D., is assistant professor of ophthalmology at Department of Ophthalmology and Visual Sciences University of Iowa, Iowa City, Iowa. Dr. Abràmoff has patent applications for automated diagnosis of diabetic retinopathy and glaucoma. He can be e-mailed at michael-abramoff@uiowa.edu.

Stephen R. Russell, M.D., is professor of ophthalmology and service director of the Vitreoretinal Diseases and Surgery department at the University of Iowa. Dr. Russell has no proprietary interests in any of the products. He can be e-mailed at steve-russell@uiowa.edu.

References

1. American Academy of Ophthalmology Retina Panel, Preferred Practice Patterns Committee. Diabetic retinopathy. 2003:1-33.

2. Klein R. The epidemiology of diabetic retinopathy: findings from the Wisconsin Epidemiologic Study of Diabetic Retinopathy. Int Ophthalmol Clin. 1987;27:230-238.

3. American Telemedicine Association. Standards and Guidelines in Telehealth for Diabetic Retinopathy: American Telemedicine Association Position Statement. Presented at: American Telemedicine Association Ninth Annual Meeting; May 2-5, 2004; Tampa, Fla.

4. Klonoff DC, Schwartz, DM. An economic analysis of interventions for diabetes. Diabetes Care. 2000;23:390-404.

5. Abramoff MD, Suttorp-Schulten MS. Web-based screening for diabetic retinopathy in a primary care population: the EyeCheck project. Telemed JE Health. 2005;11:668-674.