Telemedicine’s potential to disrupt norms paves the way for new paradigms. But how far is retina from employing the benefits it may yield?
Mrs. H wakes up at 6:00 am and gets ready for the day. Before eating breakfast, she grabs her phone and checks her email, browses the news, and scrolls through her Facebook feed. At 8:00 am she receives a reminder on her phone to test her eyes. She was diagnosed with diabetes 20 years ago, and her doctors have been monitoring her eyes, as she has needed anti-VEGF injections in the past.
She opens the testing app and first performs a home vision-monitoring test. After both eyes are tested, her results are uploaded to the cloud, and the algorithm determines any changes from baseline. There is a slight change, so the app prompts her to insert the phone into a virtual reality–like headset, which she then puts on her head. Mrs. H is instructed to look at specific targets on a beautiful mountain landscape, and, unbeknownst to her, a digital fundus photo and an optical coherence tomography (OCT) scan are obtained. These images are uploaded to the cloud, where image recognition software with deep learning capabilities identifies a change in her exam, identifying recurrent macular edema with a corresponding decrease in vision.
An alert is sent to her retina specialist, whose office calls to set up an appointment for treatment. Mrs. H has the option to come into the office during regular hours or, as part of a premium service, a doctor from the practice can come to her house to do a full examination, testing with portable equipment, and perform any necessary treatment.
Sound like something out of Star Trek? Not so fast. Many of the above technologies are now available, or will be available within 5 years. Telemedicine is the fastest growing segment of medicine, growing 50% just from 2013 to 2015.1 Approximately 15 million people used telemedicine services in 2015.1 Additionally, 29 states have laws requiring commercial insurers to cover telemedicine services, and eight more states have similar bills under discussion.1 For Medicaid patients, 87% of states cover statewide telemedicine services, a shift from previous restrictions of the technology to only rural or geographically isolated locations.1
Telemedicine is experiencing exponential growth and adoption, significantly changing the practice of medicine as we know it. Some retina specialists may think that the general trend toward telemedicine will not affect retina practice, but this is unlikely.
Telemedicine screening for diabetic retinopathy (DR) and glaucoma has been in use for more than a decade. Several large academic institutions have well-established telemedicine screening protocols. These protocols include obtaining fundus images at primary care doctors’ offices, retail pharmacies, community health centers, and other locations. Images are transferred to a cloud-based storage system, from which trained image readers can flag abnormal images for physician review. Readers are limited by time, so they have limited ability to screen large numbers of images and input data. Data are entered into software that determines the severity of DR and recommends follow-up based on current guidelines. Multiple studies have established the benefit of telemedicine screening in DR, and this capability will be important for the future as the prevalence of diabetes continues to grow.2-4
To meet the demand for this growing population affected by diabetes, sophisticated software programs, including image analysis and recognition, deep learning, and predictive analytics tools, are poised to enter the mix. Several telemedicine diabetic screening programs are beginning to use automated image analysis to screen fundus images for the presence of DR. These tools can screen thousands of images in a fraction of the time needed by human telemedicine graders.
With the advent of deep learning and neural network software, the data from thousands of fundus photos and OCT images can provide key examples of pathology or lack thereof. Feature extractors will highlight and tag key anatomic regions, boundaries, and abnormalities within fundus images and OCT volume scans. Deep learning and neural networks will then separate the data from fundus images and OCT scans into categories, such as normal; mild, moderate, and severe nonproliferative DR with or without macular edema; and proliferative DR.
Automated data analysis will then be combined with predictive analytics to further enhance the capabilities. Predictive analytics is the use of mining of historical data to develop advanced statistical models of key variables to predict future trends and behaviors. Potential insights with predictive analytics tools include identifying patients at highest risk of progression or blinding complications, identifying which patients may respond best to specific treatments and medications, and determining which patients may need closer monitoring.
Home vision monitoring will become their new norm for patients identified as at high risk for DR complications. Several companies now offer home vision monitoring, and two have been cleared by the US Food and Drug Administration.
Home vision monitoring allows users to test themselves without visiting a doctor’s office. Results are then automatically uploaded to cloud software where the data are processed. Algorithms send alerts to physicians when a patient’s vision testing results deviate from his or her norm. Home monitoring tools will be key to earlier diagnosis and improved outcomes.
Another facet of telemedicine that presents exciting opportunities and accounts for much of the current hype in medicine involves live or asynchronous patient evaluations. These may include virtual visits, virtual consultations, or virtual expert second opinions. Several technological advances now in the works must be realized before virtual retina visits are feasible, but we can expect these technologies to be in physicians’ and patients’ hands within the next 5 years.
Several innovation labs and companies are developing technologies that would allow retina fundus selfies by patients, as well as low-cost OCT technologies that could be powered by a smartphone. It is only a matter of time before a virtual-reality retina headset can obtain nonmydriatic fundus images and OCT images. The development of these technologies is the major barrier to achieving virtual retina visits; however, I anticipate these hurdles will be cleared in the next few years.
Virtual consultations are also being piloted at several hospitals. Using handheld nonmydriatic fundus cameras, images are obtained in the emergency room setting and wirelessly transferred to a secure device. Images are then securely transferred via an app so that ophthalmologists can review the findings and suggest treatment plans. Face-to-face visits between patients and physicians can then be performed via secure video links.
Virtual expert second opinions are also popular in other specialties, including neurosurgery, orthopedics, and otolaryngology, and they are starting to be piloted in ophthalmology and, specifically, retina. These services facilitate expert subspecialist second opinions, and they could be extremely useful for patients with atypical uveitis, ocular oncologic issues, or other types of retinal disease in which there is a dearth of specialists in a particular region.
DISRUPTING MEDICINE FOR THE BETTER
Technology continues to advance at lightning-fast speed. Telemedicine is disrupting many areas of medicine as legislation lowers the barriers to more widespread adoption. For retina specialists, telemedicine will play a vital role in ensuring that we can provide the best care to an ever-increasing number of patients with sight-threatening diseases.
Sophisticated telemedicine screening protocols combined with automated image analysis, deep learning, predictive analytics, and remote home monitoring will complement our in-office examinations, evaluations, and treatments. These tools will allow earlier diagnosis and treatment, leading to improved outcomes, improved quality of care, and decreased costs. n
1. American Telemedicine Association. Top 5 telemedicine trends. MedCityNews.com. Posted May 6, 2016. Accessed September 7, 2016. medcitynews.com/2016/05/top-5-telemedicine-trends/?rf=1
2. Gadkari S. Innovative model for telemedicine-based screening for diabetic retinopathy in the developing world. Can J Ophthalmol. 2016; 51(3):e109-e111.
3. Williams GA, Scott IU, Haller JA, et al. Single-field fundus photography for diabetic retinopathy screening: a report by the American Academy of Ophthalmology. Ophthalmology. 2004:111(5):1055-1062.
4. Park DW, Mansberger SL. Eye disease in patients with diabetes screened with telemedicine [published online ahead of print June 21, 2016]. Telemed J E Health.
5. Whiting DR, Guariguata L, Weil, C, Shaw J. IDF Diabetes Atlas. Global Burden: Prevalence and Projections, 2010 and 2030. Diabetes Res Clin. 2011;94(3):311-321.
6. United Nations Department of Economic and Social Affairs, Population Division. World Population Prospects. The 2015 Revision. 2015.
7. Centers for Disease Control and Prevention. National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States, 2014. Atlanta, GA: U.S. Department of Health and Human Services; 2014.
8. National Eye Institute. Diabetic Retinopathy. Accessed September 7, 2016. nei.nih.gov/eyedata/diabetic
S.K. Steven Houston III, MD
• retina surgeon, Florida Retina Institute, Orlando, Fla.; faculty, department of telemedicine and innovative technologies, Wills Eye Hospital, Philadelphia
• financial interest: consultant for Vital Arts & Sciences
• @trieyedoc; email@example.com