Retina specialists and their patients have learned how AI and telemedicine can be effective in monitoring disease progression and alerting clinicians to early signs of adverse changes.
Before the COVID-19 pandemic, retina specialists had limited experience with artificial intelligence (AI) and digital healthcare, which can include services such as telemedicine. Interactions with AI in retina were perceived as something—although feasible—best kept for future implementation.
‘Digital healthcare’ was a vague term, pregnant with promise but lacking a clear use case for daily patient interactions. However, over the past 2 years, retina specialists and their patients have learned how AI may be a tool to monitor disease progression and how telemedicine might mean more than a mere video chat between patient and clinician.
Whilst applications of digital healthcare differ from specialty to specialty, and patient expectations and appetites for such offerings may vary based on personal experience, it behoves modern retina specialists leveraging digital healthcare in their clinics to ensure that patients understand its benefits and limitations in retina.
The increasing ubiquity of technological platforms replacing analogue media—from QR codes replacing printed restaurant menus to Zoom conferences in lieu of in-person meetings—has resulted in many retina patients feeling more comfortable with the concept of digital medicine.
Even among our retired patients, whose lives may be less dictated by video conferencing than their working-age counterparts, the learning curve for digital interfacing has been flattened. This means that a general aversion to digital medicine, one of the pre-2020 challenges faced by retina specialists who wished to implement AI-based services into their care models, has largely been erased.
Still, some retina patients may have unrealistic expectations about how digital healthcare can alleviate the burden associated with retinal diseases, such as wet age-related macular degeneration (AMD). Although the point is obvious to clinicians, I sometimes remind my patients that I cannot perform a comprehensive eye examination via video conference and am unable to evaluate visual function based on subjective self-reported assessment.
For some conditions, telemedicine’s limitations mean that in-clinic examinations remain the standard for evaluation. For other conditions, digital options offer cutting-edge healthcare.
Consider the example of a patient with dry intermediate AMD. In a more analogue age, such a patient would have been handed an Amsler grid and instructions to call the clinic if the grid’s uniform pattern became distorted, as such a change might indicate a conversion from intermediate to wet AMD. However, this model could be prone to lack of adherence and a misunderstanding of instructions, and placed the burden of initiating care on the patient.
If that same intermediate AMD patient was referred to the ForeseeHome AMD Monitoring Program (Notal Vision), they would be enrolled in a modern AI-based digital healthcare framework. After a referring physician has determined that a patient with dry intermediate AMD is well suited for home monitoring, the Notal Vision Monitoring Centre, which is the provider of the aforementioned program, works with the patient to set up the device, teaches them how to perform daily preferential hyperacuity perimetry tests and serves as a troubleshooting liaison.
The monitoring centre tracks patient adherence, and contacts those who are not regularly performing at-home tests to ensure that there are no hardware problems or other healthcare issues which are interfering with consistent evaluation. Results of routine at-home testing are sent to the Notal Vision Monitoring Centre, which relies on an AI algorithm to detect aberrations in patient performance that might indicate a conversion from intermediate to wet AMD.
If AI flags the patient’s test results, an in-house ophthalmologist reviews the report and pings the referring physician’s clinic if it is determined that the referring physician should be alerted. The clinic may then reach out to the patient to schedule an in-person examination to see whether conversion to wet AMD has occurred.
In this digital healthcare model, an intermediate AMD patient receives routine AI-based monitoring services whilst outside the clinic in addition to the regularly scheduled visits that they have with their provider. Retina specialists should note that such a home monitoring program (covered by Medicare in the United States) is not a stand-in for clinical examination, but is rather a digital supplement to in-person visits.
The real-world case of a 75-year-old woman, identified as J.B., illustrates the value of home monitoring for wet AMD conversion. J.B. was a patient with intermediate AMD OD and wet AMD OS, which was treated with aflibercept (Eylea, Regeneron Pharmaceuticals). Visual acuity (VA) measured 20/30 OD and 20/150 OS at her most recent visit. Given that she was a high-risk patient with intermediate AMD and a history of wet AMD in the fellow eye, I referred J.B. to the Notal Vision Monitoring Centre for enrolment in the ForeseeHome AMD Monitoring Program.
Six months after initiating at-home monitoring, my clinic received an alert that J.B.’s most recent test flagged her for possible conversion to wet AMD OD. My clinic scheduled an in-person examination, at which point I found that her visual acuity (VA) remained stable at 20/30 OD.
Fundus photography revealed macular drusen and changes to the retinal pigment epithelium and did not show evidence of subretinal haemorrhage (Figure 1). Optical coherence tomography imaging revealed only borderline findings (Figure 2). Fluorescein angiography depicted a small choroidal neovascular membrane temporal to the optic disc that did not involve the central macula (Figure 3).
I diagnosed J.B. with wet AMD based on these findings. Initiation of aflibercept therapy resulted in stable VA of 20/25 OD.
In this instance, home-based monitoring detected asymptomatic conversion to wet AMD from intermediate AMD in J.B., whilst VA was at least 20/40, which is particularly good news when one considers the findings of a 2020 real-world study that detecting wet AMD before VA dips below 20/40 is key to maintaining 20/40 vision after 1 or 2 years of anti-vascular endothelial growth factor therapy. In the case of J.B., detection of wet AMD might have occurred after she passed the 20/40 VA threshold if not for use of the ForeseeHome AMD Monitoring Program.
Digital healthcare offerings frame one’s practice as a cutting-edge institution that prioritises the latest in retina care. Although some patients might have a different concept of what digital healthcare looks like, educating them on the parameters of telehealth vis-à-vis retina care will set reasonable expectations.
Furthermore, implementation of digital healthcare platforms, such as the ForeseeHome AMD Monitoring Program, may allow practices to leverage the power of AI to offer care outside the bricks-and-mortar structure of the clinic, thereby launching practices into the future of medicine.