Eyeworld

SUMMER 2026 | EYEWORLD | 71 G Relevant disclosures Wang: None Wirostko: Icare USA, Injectsense, MyEyes, Qlaris Bio Contact Wang: sywang@stanford.edu Wirostko: barbara.wirostko@hsc.utah.edu headset to a patient's home and collect visual fields monthly, which will enable physicians to identify patients with progression earlier. While clinician informaticists like Dr. Wang and Dr. Wirostko are excited about the questions they hope to answer in the future thanks to digital tools and big data, there are still hurdles to overcome. One of the main hurdles is that right now much of the information is stored in different systems. "Dr. Wirostko has her mountain of home tonometry data, and we've got our EHR. Being able to combine those different kinds of data and come up with insights across multiple mo- dalities is a challenge," Dr. Wang said. "If we're looking at imaging, too, there are different kinds of imaging, different vendors, different systems for storing the imaging. How we pull all of that together in a standardized way to really enable reproducible studies I think is a chal- lenge in our field." A word of caution from Dr. Wang regarding the advances that big data and AI might bring to the glaucoma field: If you're going to consider deploying an algorithm into your practice, keep in mind the generalizability of that algorithm. How was the algorithm created? What datasets was it tested in? Will it work in your patient population? "You don't want to accidentally deploy a tool that is going to amplify health disparities in some way," she said. "Keep these things in mind as we enter this brave new world of AI in clinical care. We should be cautious and do our due diligence, asking for these details on any clinical AI algorithm we are considering." When it comes to adopting newer digital technologies for glaucoma—now or in the fu- ture—Dr. Wirostko said there will be a learning curve, and she sees glaucoma specialists as paving the way for such technologies' use on a broader scale. "We are always looking at new and better ways to diagnose and prevent disease. There's nothing worse than seeing a patient you've been following and managing for 10 years continue to lose vision, especially with relatively good IOPs," Dr. Wirostko said. "This is where I would love to see big data help determine other risk factors that we may not be considering." A patient story in digital, data-driven glaucoma care Dr. Wirostko offered a specific patient exam- ple that highlighted how new data-driven tools are altering patient care for glaucoma and highlighting areas of new research. She had a patient from Cabo San Lucas, Mexico, with pigmentary glaucoma. This patient had the iCare HOME2 and a virtual reality visual field headset. The monthly visual field data suggested he was progress- ing. Dr. Wirostko said she had him getting up earlier and earlier in the morning to check his IOP, and this revealed that, despite having PRESERFLO and pressures that were generally between 10–11 mm Hg in waking hours, in the early morning hours he was spiking to 17–18 mm Hg. "For his advanced damage, this was too high for his nerve," she said. At this point, the patient flew up to see Dr. Wirostko and her colleagues in person. They advised he would likely need a tube shunt or trabeculectomy, which the patient preferred to have performed in Atlanta, Georgia, where he had an apartment to stay in during postop care. Between flying back to Cabo and awaiting his surgery in Atlanta, the patient learned he had sleep apnea from another physician. "He stopped breathing about 44 times an hour," Dr. Wirostko said, noting he didn't have typical physical risk factors for sleep apnea. With CPAP treatment for sleep ap- nea, Dr. Wirostko said his early morning IOP fluctuations completely flattened. "We always thought that sleep apnea had an association with glaucoma and that the association was due to low oxygen levels or hypoxia to the optic nerve," Dr. Wirostko said. "But now we're wondering, what if sleep apnea is actually responsible for some of these pressure spikes? That's still un- known, and more research has to be done, but as we continue to collect data and think outside the box, I think we are going to an- swer more of these questions. … But again, these are just hypotheses based on individu- al cases, which is why we need big data."

• Blank Page
• Blank Page (1)