Clinical outcomes in neovascular age-related macular degeneration: a cohort study of patients with care delay due to the COVID-19 pandemic

Abstract The COVID-19 pandemic has led to both intentional and unintentional care delay among age-related neovascular macular degeneration (nvAMD) patients. Prior studies have demonstrated that patients who discontinue nvAMD treatment for prolonged intervals are at high risk for vision loss, but less is known regarding shorter-term delay, such as during the height of the pandemic. Previous studies have looked at COVID-19 related delay in care and have shown a loss of visual acuity (VA) among these patients, but studies are limited by short follow-up or insufficient comparisons. This was an observational cohort study of nvAMD patients from March 1, 2019, through July 1, 2021, who experienced care delay. VA was modeled using a linear longitudinal mixed-effects model comparing historic data pre-lockdown to data post-lockdown. Covariates included baseline anatomic variables, demographic variables, and time intervals (treatment interval, delay interval). Secondary anatomic and treatment outcomes were modeled using a multilevel binary logistic regression model. 163 eyes among 116 patients were included. Initial longitudinal mixed-effects models found that although overall VA decreased at a yearly rate, when comparing pre-lockdown and post-lockdown time periods, VA slopes were not statistically different. Single-covariate longitudinal models showed that age, sex, and delay interval significantly affected VA slope. The multivariate longitudinal model found that a longer delay interval significantly decreased rate of VA loss. Multilevel binary logistic regression models showed a significant increase in odds of anti-VEGF treatment, presence of subretinal fluid, and macular hemorrhages in the post-lockdown period. Overall, when compared to historic data, rate of VA loss among our cohort did not vary significantly in pre-versus post-lockdown time periods, although treatment and anatomic variables did worsen post-lockdown suggesting that patients may be appropriately delayed but this comes at the risk of increased need for treatment