How Can Big Data Help to Triage Ophthalmology Patients?

New study results published in JAMA Ophthalmology suggest researchers can leverage big data to triage ophthalmic clinic appointments, balancing the glaucoma progression risk against the morbidity risk from coronavirus disease 2019 (COVID-19) exposure during ophthalmic care.

Due to social distancing measures currently in place across the country, eye care professionals have had to determine whether patients’ appointments can be safely postponed. This process involves weighing the risks of patients’ glaucoma worsening during the interim vs the morbidity risk of acquiring COVID-19 while seeking ophthalmic care. When social distancing restrictions are eased, ophthalmologists will also need to prioritize which patients they see first.

Patients with chronic ocular diseases, such as glaucoma, risk experiencing irreversible visual acuity or visual field loss from disease progression if they are unable to receive timely medical care, the researchers explain. “Yet a disproportionately large number of patients with glaucoma are also of older age and have medical comorbidities that predispose them to morbidity and mortality from COVID-19.”

In a cross-sectional study, the researchers used information from a large data repository to develop a flexible and scalable scoring algorithm for patients with glaucoma to aid in triaging those seeking ophthalmologic care.

A total of 1034 patients enrolled in the Sight Outcomes Research Collaborative (SOURCE) Ophthalmology Electronic Health Record Data Repository were included in the study. Specifically, all patients over the age of 18 receiving care at the University of Michigan Kellogg Eye Center, Ann Arbor, between March 16, 2020, and April 16, 2020, were identified.

The researchers classified old age (over 65 years), pregnant or breastfeeding women, and those with chronic morbidities as high-risk populations for COVID-19. Greater statistical weight was also given to those over the age of 80 compared with those aged 65 to 79 years.

Glaucoma severity was characterized based on incisional intraocular surgery in the past 3 months, records of high (≥30 mm Hg) or low (<6 mm Hg) intraocular pressure in the past year, considerable visual field loss in 1 or both eyes, and monocularity.

“For each patient, a COVID-19 morbidity risk score (CS) and glaucoma severity and progression risk score (GS) were determined by assigning point values to the noted characteristics and calculating the sum,” the authors write.

Of those included in the study, the majority were women (56.5%), and the average age was around 67.

The algorithm determined:

• Average (SD) glaucoma severity and progression risk score was 4.0 (14.4) points, average COVID-19 morbidity risk score was 27.2 (16.1) points, and average total score (TS) was 31.2 (21.4) points
• During pandemic-associated reductions in services, using total score thresholds of 0, 25, and 50 points, 970 appointments (93.8%), 668 appointments (64.6%), and 275 appointments (26.6%), respectively, were identified for postponement and rescheduling

The researchers also demonstrated how the algorithm could be used for specific cases, applying the scoring system to 3 patients.

The scoring algorithm was then validated against glaucoma specialists’ review of patients’ electronic health records to assign them a tier for follow-up care. The findings of the validation exercise “demonstrate that the TSs from our algorithm aligned with glaucoma specialists’ preferences for appointment triaging for most patients,” the authors write.

However, short-term complications and long-term sequelae from COVID-19 were not included in the analysis, marking a limitation to the study. Additional research is also needed to confirm whether the researchers’ prioritization of clinic appointments, based in part on the scoring algorithm, resulted in improved outcomes.

“While our approach is tailored to patients with glaucoma, we believe similar methods may be developed to help offer guidance for appointment rescheduling for other ophthalmic and nonophthalmic conditions during this pandemic and similar situations in the future,” the authors conclude.

Reference

Bommakanti NK, Zhou Y, Ehrlich JR, et al. Application of the sigh outcomes research collaborative ophthalmology data repository for triaging patients with glaucoma and clinic appointments during pandemics such as COIVD-19. JAMA Ophthalmol. Published online July 17, 2020. doi:10.1001/jamaophthalmol.2020.2974