In light of COVID-19, researchers recently explored ways clinicians can best prioritize glaucoma patients who need to be seen and who can wait—and they found artificial intelligence can help.

Researchers from the University of Michigan analyzed data from a large electronic health record repository to create a scoring system that considered a patient’s glaucoma severity and their risk of experiencing disease progression against their risk of experiencing morbidity from COVID-19. 

“Since the COVID pandemic began, ophthalmologists and optometrists have been placed in the difficult position of trying to decide which of our patients’ clinic appointments should get postponed and rescheduled and which ones should be kept, carefully considering the benefits and risks of their coming in to clinic to see us,” says researcher Joshua D. Stein, MD.

 For patients with sight-threatening conditions such as glaucoma, this can be a particularly difficult decision because delays in care can result in progression that can lead to irreversible blindness, he adds. Yet, on the flip side, many patients with glaucoma are also elderly individuals who have other medical comorbidities, and if they are exposed to COVID-19 as a result of seeking eye care services, the risk of morbidity and mortality can be high, Dr. Stein says. 

The study’s automated system generated a score that captures these competing risks for each patient, he explains.

“Ophthalmologists and optometrists can use these scores to help them decide which of their patients would be better off having their clinic appointments kept vs. others where postponing and rescheduling the appointment would be the safer option,” he says.

Furthermore, as federal, state and local restrictions are lifted and it is safer for patients to come in to seek eye care, clinicians can also make use of the algorithm scores to prioritize which patients are in greater need for follow-up care ahead of others, making it easier to identify the patients who would most benefit from coming in to the clinic, Dr. Stein adds. 

The study identified 1,034 patients with upcoming glaucoma clinic appointments from March to April 2020 who were part of the Sight Outcomes Research Collaborative Ophthalmology Electronic Health Record Data Repository. They developed a risk stratification tool that calculated a glaucoma severity and progression risk score and a COVID-19 morbidity risk score that were added together for a total score for each patient.

Subjects were approximately 67 years old, and more than half were women (55.6%) and white (71%). The mean glaucoma severity and progression risk score was 4 points, while the mean COVID-19 morbidity risk score was 27.2 points, for a mean total score of 31.2 points.

The study assessed different triage thresholds (zero, 25 and 50) to demonstrate varying degrees of caution regarding the adverse outcomes, whether related to COVID-19 exposure or the risk of glaucoma progression.  In this sample, a threshold of zero meant 93.8% of the scheduled appointments could safely be postponing for up to three months. The total score thresholds of 25 and 50 points suggested 64.6% and 26.6% of appointments could safely be postponed, respectively.

The study also validated the total scores from the algorithm against glaucoma specialists’ recommendations based on the review of patients’ records and found the two aligned.

Bommakanti NK, Zhou Y, Ehrlich JR, et al. Application of the sight outcomes research collaborative ophthalmology data repository for triaging patients with glaucoma and clinic appointments during pandemics such as COVID-19. JAMA Ophthalmology. July 17, 2020. [Epub ahead of print].