How AI can potentially close a care gap for people with diabetes

Michael Abramoff, MD, PhD -

As an ophthalmologist, I see too many individuals with diabetes who lose vision needlessly from diabetic retinopathy, primarily because they are not diagnosed and treated in time.

Thankfully, new advances in technology, particularly in AI, promise to close this care gap and enable early detection where these individuals are, in primary care.

Diabetic retinopathy (DR) is the leading cause of blindness among working age adults in the United States. The disease causes vision loss in an estimated 24,000 Americans annually. However, in 90% of people with diabetes, blindness can be prevented or delayed through regular retinal examinations and treatment, according to the U.S. Centers for Disease Control (CDC).

That’s why so many eyecare professionals find it frustrating that DR continues to plague the U.S. healthcare system year after year. Simply put: We are aware that DR is a problem and we know how to manage and treat it, yet in recent decades we’ve made little progress reversing the trend – in large part because as many as half of all people with diabetes do not receive recommended annual eye exams, according to the CDC.

More alarmingly, without a change in our approach to detecting DR, the problem could get worse because of the skyrocketing rate of diabetes across the U.S. The data are worrisome and eye-popping: Between 2017 and 2045, the number of people worldwide with diabetes is expected to increase 48 percent to 629 million, according to the International Diabetes Federation.

It doesn’t have to be this way - and it shouldn’t be this way. Placing autonomous artificial intelligence-based diagnostics for DR in PCP offices may increase the number of people with diabetes who get their retinal exam - because that is where individuals with diabetes receive their care most often.

Understanding the scale of diabetic retinopathy
To understand the potential benefits of AI-based diagnostics for DR in primary care, it is important to first understand the scale of the problem and how it’s currently being addressed.

Every one of the more than 30 million people with diabetes is at risk of developing DR. About 33 percent of people with diabetes have some degree of DR and about 10 percent will develop a vision-threatening form of the disease, making the condition the leading cause of vision loss in working-age adults, according to the International Agency for the Prevention of Blindness.

Among the most pernicious aspects of DR is that it is initially asymptomatic – meaning that most people with diabetes have no idea that they have begun developing the disease. Only when the condition enters its late stages do people with diabetes develop DR symptoms and realize they have been affected, but by then it is often too late for treatment and some permanent damage has occurred.

Diabetes care typically happens in primary care, where both diabetes itself and its potential complications can be dealt with, except for the retinal exam. Because of its complexity, it is routed to an eye care provider, which includes ophthalmologists and optometrists. For example, during an appointment with a person who has diabetes, a PCP may refer the patient to an eye care provider, who then examines the retina at the back of the eye for damage.

Shortcomings of the current approach to DR treatment
Though referrals from PCPs to eye care providers are common, many people with diabetes find it difficult to complete the exam due to the time, cost, and scheduling challenges associated with a specialist visit. Someone who is asymptomatic, may view overcoming these obstacles as especially burdensome. In fact, less than half of all individuals with diabetes in the U.S. do not undergo their recommended annual eye exams.

The current process can be arduous, complicated and demotivating for patients. Contrast that scenario with completing the exam during a visit to an individual’s PCP, who is also the medical professional the patient sees regularly for diabetes care and with whom he or she is likely most comfortable.

A shortage of specialists
Even if exam compliance was not an issue, there may not be enough eye care specialists in the U.S. to administer annual retinal exams to our country’s 30 million (and growing) people with diabetes. That number dwarves the nation’s current supply of approximately 20,000 ophthalmologists and 36,000 optometrists, according to figures from the U.S. Bureau of Labor Statistics and American Academy of Ophthalmology.

Looking toward an automated future
Even though the case for testing for DR in primary care is strong, many PCPs do not currently feel comfortable performing a retinal exam. However, with the FDA’s recent authorization of IDx-DR, the first autonomous AI diagnostic system authorized to detect DR, that may change.

IDx-DR leverages AI-based diagnostics to enable PCPs to perform the dilated retinal exam and interpretation in their clinics. The IDx-DR system is indicated for use with the Topcon NW400, an automated fundus camera that anyone with a high school degree can be trained to properly use. The system provides instantaneous results right at the point of care in the PCP’s office, potentially reducing the need for unnecessary referrals. If the technology detects more than mild DR, or diabetic macular edema the patient can then be referred to a specialist.

This technology may have a strong appeal to PCPs because it enables them to provide more comprehensive diabetes care and ensure patients are tested for DR.

By reducing the need for human evaluation, automated technology may provide convenient access to DR testing, which in turn has the potential to lead to a reduction in one of the world’s leading causes of preventable blindness.

Dr. Michael D Abramoff, MD, PhD, is President and Founder of IDx. Abramoff is a fellowship‐trained retina specialist with a PhD in image analysis. He is the Robert C. Watzke Professor of Ophthalmology and Visual Sciences at the University of Iowa, with joint appointments at Biomedical Engineering, and Electrical and Computer Engineering.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.