As AI eye exams prove their worth, lessons for future tech emerge.

Christian Espinoza, director of a Southern California drug treatment provider, recently began employing a powerful new assistant: an artificial intelligence algorithm that can perform eye exams with images taken from a retina camera. It provides quick diagnosis without the presence of a doctor.

His clinic, Tarzana Treatment Centers, is among the early adopters of an AI-based system that promises to dramatically increase screening for diabetic retinopathy, the leading cause of blindness in working adults and An estimated 38 million Americans are at risk. Diabetes

“This is a gift for us,” said Espinoza, the organization’s director of clinic operations, citing the benefits of a quick and easy screening that can be administered with little training and provides immediate results.

His patients love him too. Joseph Smith, who has type 2 diabetes, recalled the arduous task of taking a bus to the eye specialist, having his eyes dilated, and then waiting a week for the results. “It was scary,” he said. “Now, it takes minutes.”

Amid all the buzz about artificial intelligence in healthcare, eye exam technology is emerging as one of the first proven use cases for AI-based diagnostics in a clinical setting. While the FDA has approved hundreds of AI medical devices, adoption has slowed as vendors navigate the regulatory process, insurance coverage, technical hurdles, equity concerns, and the challenges of integrating them into provider systems. have been.

Eye exams show that AI’s ability to provide quick results, as well as cost savings and the convenience of not needing an additional appointment, have major benefits for both patients and providers. Of the roughly 700 eye exams performed at Espinoza’s clinics over the past year, retinopathy was detected in about a quarter, and patients were referred to a specialist for further care.

Diabetic retinopathy results when high blood sugar damages the blood vessels in the retina. While managing a patient’s diabetes can often prevent the disease — and there are treatments for more advanced stages — doctors say regular screenings are crucial to catching symptoms early. An estimated 9.6 million people in the United States suffer from this disease.

Three companies with FDA-approved AI eye exams for diabetic retinopathy—Digital Diagnostics, based in Coralville, Iowa; Ianke of Woodland Hills, Calif.; And Israeli software company AEYE Health has sold systems to hundreds of practices nationwide. A few dozen companies have conducted research in the narrow field, and some have regulatory clearance in other countries, including tech giants like Google.

Digital Diagnostics, formerly Idx, won FDA approval for its system in 2018, after decades of research and a clinical trial involving 900 patients diagnosed with diabetes. It was the first fully autonomous AI system in any field of medicine, making its approval “a landmark moment in medical history,” said Aaron Lee, a retina specialist and an associate professor at the University of Washington.

John Bertrand said the system used by Tarzana Treatment Centers can be operated with a high school degree and a few hours of training, and takes only minutes to diagnose, mostly without eye dilation. , CEO of Digital Assessment.

The setup can be placed in any dimly lit room, and patients stare at the camera with their face resting on their chin and forehead while a technician takes pictures of each eye.

The American Diabetes Association recommends that people with type 2 diabetes be screened every one to two years, yet only 60 percent of people living with diabetes get an annual eye exam, said Robert Gabe, ADA’s chief scientific and medical officer. Inspect. Rates can be as low as 35 percent for people 21 and younger with diabetes.

In many parts of the United States, a shortage of ophthalmologists and ophthalmologists can make scheduling appointments difficult, sometimes with bookings for months. In addition, the barriers to traveling to an additional appointment to dilate their eyes—which means taking time off work or school and securing transportation—can be especially difficult for low-income patients. They are also at a higher risk of type 2 diabetes.

“Ninety percent of our patients are blue-collar,” said Espinoza of his Southern California clinic, which serves a largely minority population. “If they don’t work, they don’t eat.”

A potential downside to not having a doctor screen, Lee said, is that the algorithm looks solely for diabetic retinopathy, so it may miss other diseases like choroidal melanoma. Algorithms also typically “err on the side of caution” and over-refer patients.

But the technology has shown another big advantage: According to a recent Stanford University study, follow-up after a positive result is three times higher with an AI system.

That’s because of “the immediacy of the message,” said David Myung, an associate professor of ophthalmology at the Byers Eye Institute at Stanford. When it is delivered immediately, rather than weeks or months later, it is more likely to be heard and acted upon by the patient.

Myung launched Stanford’s automated teleophthalmology program in 2020, which originally focused on telemedicine and then transitioned to AI at his Bay Area clinic. That same year, the National Committee for Quality Assurance expanded screening criteria for diabetic retinopathy to include AI systems.

Meung said it took Stanford Health System’s cybersecurity and IT systems about a year to integrate the new technology. Myung said there was also a learning curve, especially for taking quality images that the AI ​​could understand.

“Despite hitting our stride, there’s always something to improve,” he added.

The AI ​​test is powered by the Centers for Medicare and Medicaid Services’ reimbursement code, which can be difficult and time-consuming to obtain for successful devices. But health care providers need government approval to receive reimbursement.

In 2021, CMS has set the national payment rate for AI diabetic retinopathy screening at $45.36, well below the median privately negotiated rate of $127.81, according to a recent report. New England Journal of Medicine The study of AI. Each company’s business model is slightly different, but they typically charge providers a subscription or licensing fee for their software.

The companies declined to share what they charge for their software. Cameras can cost up to $20,000 and are either purchased separately or as a rental bundled into a software subscription.

The machines enable greater compliance with screening recommendations, said Lindsey Buchholz, clinical informatics lead at Nebraska Medicine, which began using Eyenuk’s system in mid-December. With the same addition, it makes it worthwhile.

“It kind of helps the camera pay for itself,” he said.

According to the company, today the digital assessment system is in place at about 600 sites across the country. AEYE Health said its eye exam is used by “less than hundreds” of U.S. providers. Inik declined to give details about his access.

The technology is moving forward with clinical studies for additional cameras — including a handheld imager that can screen patients in the field — and looking at other eye diseases such as glaucoma. Innovations have characterized radiology, cardiology, and dermatology, as well as ophthalmology, where AI innovation is accelerating.

“They’re going to come out in the near future — cameras that you can use in street medicine — and that’s going to help a lot of people,” Espinoza said.

2024 KFF Health News. Distributed by Tribune Content Agency, LLC.